id sid tid token lemma pos web-archive-org-7310 1 1 A a DT web-archive-org-7310 1 2 Large large JJ web-archive-org-7310 1 3 - - HYPH web-archive-org-7310 1 4 Scale Scale NNP web-archive-org-7310 1 5 Study Study NNP web-archive-org-7310 1 6 of of IN web-archive-org-7310 1 7 the the DT web-archive-org-7310 1 8 Evolution Evolution NNP web-archive-org-7310 1 9 of of IN web-archive-org-7310 1 10 Web Web NNP web-archive-org-7310 1 11 Pages Pages NNP web-archive-org-7310 1 12 success success NN web-archive-org-7310 1 13 fail fail VB web-archive-org-7310 1 14 Nov Nov NNP web-archive-org-7310 1 15 JUL JUL NNP web-archive-org-7310 1 16 Sep Sep NNP web-archive-org-7310 1 17 09 09 CD web-archive-org-7310 1 18 2009 2009 CD web-archive-org-7310 1 19 2011 2011 CD web-archive-org-7310 1 20 2012 2012 CD web-archive-org-7310 1 21 109 109 CD web-archive-org-7310 1 22 captures capture NNS web-archive-org-7310 1 23 23 23 CD web-archive-org-7310 1 24 Jun Jun NNP web-archive-org-7310 1 25 2003 2003 CD web-archive-org-7310 1 26 - - SYM web-archive-org-7310 1 27 15 15 CD web-archive-org-7310 1 28 Apr Apr NNP web-archive-org-7310 1 29 2021 2021 CD web-archive-org-7310 1 30 About about IN web-archive-org-7310 1 31 this this DT web-archive-org-7310 1 32 capture capture NN web-archive-org-7310 1 33 COLLECTED collect VBN web-archive-org-7310 1 34 BY by IN web-archive-org-7310 1 35 Organization Organization NNP web-archive-org-7310 1 36 : : : web-archive-org-7310 1 37 Internet internet NN web-archive-org-7310 1 38 Archive Archive NNP web-archive-org-7310 1 39 The the DT web-archive-org-7310 1 40 Internet Internet NNP web-archive-org-7310 1 41 Archive archive JJ web-archive-org-7310 1 42 discovers discover NNS web-archive-org-7310 1 43 and and CC web-archive-org-7310 1 44 captures capture NNS web-archive-org-7310 1 45 web web NN web-archive-org-7310 1 46 pages page NNS web-archive-org-7310 1 47 through through IN web-archive-org-7310 1 48 many many JJ web-archive-org-7310 1 49 different different JJ web-archive-org-7310 1 50 web web NN web-archive-org-7310 1 51 crawls crawl NNS web-archive-org-7310 1 52 . . . web-archive-org-7310 2 1 At at IN web-archive-org-7310 2 2 any any DT web-archive-org-7310 2 3 given give VBN web-archive-org-7310 2 4 time time NN web-archive-org-7310 2 5 several several JJ web-archive-org-7310 2 6 distinct distinct JJ web-archive-org-7310 2 7 crawls crawl NNS web-archive-org-7310 2 8 are be VBP web-archive-org-7310 2 9 running run VBG web-archive-org-7310 2 10 , , , web-archive-org-7310 2 11 some some DT web-archive-org-7310 2 12 for for IN web-archive-org-7310 2 13 months month NNS web-archive-org-7310 2 14 , , , web-archive-org-7310 2 15 and and CC web-archive-org-7310 2 16 some some DT web-archive-org-7310 2 17 every every DT web-archive-org-7310 2 18 day day NN web-archive-org-7310 2 19 or or CC web-archive-org-7310 2 20 longer long JJR web-archive-org-7310 2 21 . . . web-archive-org-7310 3 1 View view VB web-archive-org-7310 3 2 the the DT web-archive-org-7310 3 3 web web NN web-archive-org-7310 3 4 archive archive NN web-archive-org-7310 3 5 through through IN web-archive-org-7310 3 6 the the DT web-archive-org-7310 3 7 Wayback Wayback NNP web-archive-org-7310 3 8 Machine Machine NNP web-archive-org-7310 3 9 . . . web-archive-org-7310 4 1 Collection collection NN web-archive-org-7310 4 2 : : : web-archive-org-7310 4 3 Wikipedia Wikipedia NNP web-archive-org-7310 4 4 Outlinks Outlinks NNP web-archive-org-7310 4 5 July July NNP web-archive-org-7310 4 6 2011 2011 CD web-archive-org-7310 4 7 Crawl Crawl NNP web-archive-org-7310 4 8 of of IN web-archive-org-7310 4 9 outlinks outlink NNS web-archive-org-7310 4 10 from from IN web-archive-org-7310 4 11 wikipedia.org wikipedia.org NNP web-archive-org-7310 4 12 started start VBD web-archive-org-7310 4 13 July July NNP web-archive-org-7310 4 14 , , , web-archive-org-7310 4 15 2011 2011 CD web-archive-org-7310 4 16 . . . web-archive-org-7310 5 1 These these DT web-archive-org-7310 5 2 files file NNS web-archive-org-7310 5 3 are be VBP web-archive-org-7310 5 4 currently currently RB web-archive-org-7310 5 5 not not RB web-archive-org-7310 5 6 publicly publicly RB web-archive-org-7310 5 7 accessible accessible JJ web-archive-org-7310 5 8 . . . web-archive-org-7310 6 1 TIMESTAMPS timestamps VB web-archive-org-7310 6 2 The the DT web-archive-org-7310 6 3 Wayback Wayback NNP web-archive-org-7310 6 4 Machine Machine NNP web-archive-org-7310 6 5 - - HYPH web-archive-org-7310 6 6 https://web.archive.org/web/20110709175020/http://www2003.org/cdrom/papers/refereed/p097/P97%20sources/p97-fetterly.html https://web.archive.org/web/20110709175020/http://www2003.org/cdrom/papers/refereed/p097/P97%20sources/p97-fetterly.html NNP web-archive-org-7310 6 7 A a DT web-archive-org-7310 6 8 Large large JJ web-archive-org-7310 6 9 - - HYPH web-archive-org-7310 6 10 Scale Scale NNP web-archive-org-7310 6 11 Study Study NNP web-archive-org-7310 6 12 of of IN web-archive-org-7310 6 13 the the DT web-archive-org-7310 6 14 Evolution Evolution NNP web-archive-org-7310 6 15 of of IN web-archive-org-7310 6 16 Web Web NNP web-archive-org-7310 6 17 Pages Pages NNP web-archive-org-7310 6 18 Dennis Dennis NNP web-archive-org-7310 6 19 Fetterly Fetterly NNP web-archive-org-7310 6 20 Hewlett Hewlett NNP web-archive-org-7310 6 21 Packard Packard NNP web-archive-org-7310 6 22 Labs Labs NNP web-archive-org-7310 6 23 1501 1501 CD web-archive-org-7310 6 24 Page Page NNP web-archive-org-7310 6 25 Mill Mill NNP web-archive-org-7310 6 26 Road Road NNP web-archive-org-7310 6 27 Palo Palo NNP web-archive-org-7310 6 28 Alto Alto NNP web-archive-org-7310 6 29 , , , web-archive-org-7310 6 30 CA CA NNP web-archive-org-7310 6 31 94304 94304 CD web-archive-org-7310 6 32 dennis.fetterly dennis.fetterly RB web-archive-org-7310 6 33 @ @ NFP web-archive-org-7310 6 34 hp.com hp.com ADD web-archive-org-7310 6 35 Mark Mark NNP web-archive-org-7310 6 36 Manasse Manasse NNP web-archive-org-7310 6 37 Microsoft Microsoft NNP web-archive-org-7310 6 38 Research Research NNP web-archive-org-7310 6 39 1065 1065 CD web-archive-org-7310 6 40 La La NNP web-archive-org-7310 6 41 Avenida Avenida NNP web-archive-org-7310 6 42 Mountain Mountain NNP web-archive-org-7310 6 43 View View NNP web-archive-org-7310 6 44 , , , web-archive-org-7310 6 45 CA CA NNP web-archive-org-7310 6 46 94043 94043 CD web-archive-org-7310 6 47 manasse manasse NN web-archive-org-7310 6 48 @ @ IN web-archive-org-7310 6 49 microsoft.com microsoft.com ADD web-archive-org-7310 6 50 Marc Marc NNP web-archive-org-7310 6 51 Najork Najork NNP web-archive-org-7310 6 52 Microsoft Microsoft NNP web-archive-org-7310 6 53 Research Research NNP web-archive-org-7310 6 54 1065 1065 CD web-archive-org-7310 6 55 La La NNP web-archive-org-7310 6 56 Avenida Avenida NNP web-archive-org-7310 6 57 Mountain Mountain NNP web-archive-org-7310 6 58 View View NNP web-archive-org-7310 6 59 , , , web-archive-org-7310 6 60 CA CA NNP web-archive-org-7310 6 61 94043 94043 CD web-archive-org-7310 6 62 najork najork NN web-archive-org-7310 6 63 @ @ NFP web-archive-org-7310 6 64 microsoft.com microsoft.com ADD web-archive-org-7310 6 65 Janet Janet NNP web-archive-org-7310 6 66 L. L. NNP web-archive-org-7310 6 67 Wiener Wiener NNP web-archive-org-7310 6 68 Hewlett Hewlett NNP web-archive-org-7310 6 69 Packard Packard NNP web-archive-org-7310 6 70 Labs Labs NNP web-archive-org-7310 6 71 1501 1501 CD web-archive-org-7310 6 72 Page Page NNP web-archive-org-7310 6 73 Mill Mill NNP web-archive-org-7310 6 74 Road Road NNP web-archive-org-7310 6 75 Palo Palo NNP web-archive-org-7310 6 76 Alto Alto NNP web-archive-org-7310 6 77 , , , web-archive-org-7310 6 78 CA CA NNP web-archive-org-7310 6 79 94304 94304 CD web-archive-org-7310 6 80 janet.wiener janet.wiener NN web-archive-org-7310 6 81 @ @ NFP web-archive-org-7310 6 82 hp.com hp.com ADD web-archive-org-7310 6 83 Copyright copyright NN web-archive-org-7310 6 84 is be VBZ web-archive-org-7310 6 85 held hold VBN web-archive-org-7310 6 86 by by IN web-archive-org-7310 6 87 the the DT web-archive-org-7310 6 88 author author NN web-archive-org-7310 6 89 / / , web-archive-org-7310 6 90 owner(s owner(s NN web-archive-org-7310 6 91 ) ) -RRB- web-archive-org-7310 6 92 . . . web-archive-org-7310 7 1 WWW2003 WWW2003 NNP web-archive-org-7310 7 2 , , , web-archive-org-7310 7 3 May May NNP web-archive-org-7310 7 4 20 20 CD web-archive-org-7310 7 5 - - SYM web-archive-org-7310 7 6 24 24 CD web-archive-org-7310 7 7 , , , web-archive-org-7310 7 8 2003 2003 CD web-archive-org-7310 7 9 , , , web-archive-org-7310 7 10 Budapest Budapest NNP web-archive-org-7310 7 11 , , , web-archive-org-7310 7 12 Hungary Hungary NNP web-archive-org-7310 7 13 . . . web-archive-org-7310 8 1 ACM ACM NNP web-archive-org-7310 8 2 1 1 CD web-archive-org-7310 8 3 - - HYPH web-archive-org-7310 8 4 58113 58113 CD web-archive-org-7310 8 5 - - HYPH web-archive-org-7310 8 6 680 680 CD web-archive-org-7310 8 7 - - HYPH web-archive-org-7310 8 8 3/03/0005 3/03/0005 CD web-archive-org-7310 8 9 . . . web-archive-org-7310 9 1 Abstract Abstract NNP web-archive-org-7310 9 2 How how WRB web-archive-org-7310 9 3 fast fast RB web-archive-org-7310 9 4 does do VBZ web-archive-org-7310 9 5 the the DT web-archive-org-7310 9 6 web web NN web-archive-org-7310 9 7 change change VB web-archive-org-7310 9 8 ? ? . web-archive-org-7310 10 1 Does do VBZ web-archive-org-7310 10 2 most most JJS web-archive-org-7310 10 3 of of IN web-archive-org-7310 10 4 the the DT web-archive-org-7310 10 5 content content NN web-archive-org-7310 10 6 remain remain VBP web-archive-org-7310 10 7 unchanged unchanged JJ web-archive-org-7310 10 8 once once IN web-archive-org-7310 10 9 it -PRON- PRP web-archive-org-7310 10 10 has have VBZ web-archive-org-7310 10 11 been be VBN web-archive-org-7310 10 12 authored author VBN web-archive-org-7310 10 13 , , , web-archive-org-7310 10 14 or or CC web-archive-org-7310 10 15 are be VBP web-archive-org-7310 10 16 the the DT web-archive-org-7310 10 17 documents document NNS web-archive-org-7310 10 18 continuously continuously RB web-archive-org-7310 10 19 updated update VBN web-archive-org-7310 10 20 ? ? . web-archive-org-7310 11 1 Do do VBP web-archive-org-7310 11 2 pages page NNS web-archive-org-7310 11 3 change change VB web-archive-org-7310 11 4 a a DT web-archive-org-7310 11 5 little little JJ web-archive-org-7310 11 6 or or CC web-archive-org-7310 11 7 a a DT web-archive-org-7310 11 8 lot lot NN web-archive-org-7310 11 9 ? ? . web-archive-org-7310 12 1 Is be VBZ web-archive-org-7310 12 2 the the DT web-archive-org-7310 12 3 extent extent NN web-archive-org-7310 12 4 of of IN web-archive-org-7310 12 5 change change NN web-archive-org-7310 12 6 correlated correlate VBD web-archive-org-7310 12 7 to to IN web-archive-org-7310 12 8 any any DT web-archive-org-7310 12 9 other other JJ web-archive-org-7310 12 10 property property NN web-archive-org-7310 12 11 of of IN web-archive-org-7310 12 12 the the DT web-archive-org-7310 12 13 page page NN web-archive-org-7310 12 14 ? ? . web-archive-org-7310 13 1 All all DT web-archive-org-7310 13 2 of of IN web-archive-org-7310 13 3 these these DT web-archive-org-7310 13 4 questions question NNS web-archive-org-7310 13 5 are be VBP web-archive-org-7310 13 6 of of IN web-archive-org-7310 13 7 interest interest NN web-archive-org-7310 13 8 to to IN web-archive-org-7310 13 9 those those DT web-archive-org-7310 13 10 who who WP web-archive-org-7310 13 11 mine mine VBP web-archive-org-7310 13 12 the the DT web-archive-org-7310 13 13 web web NN web-archive-org-7310 13 14 , , , web-archive-org-7310 13 15 including include VBG web-archive-org-7310 13 16 all all PDT web-archive-org-7310 13 17 the the DT web-archive-org-7310 13 18 popular popular JJ web-archive-org-7310 13 19 search search NN web-archive-org-7310 13 20 engines engine NNS web-archive-org-7310 13 21 , , , web-archive-org-7310 13 22 but but CC web-archive-org-7310 13 23 few few JJ web-archive-org-7310 13 24 studies study NNS web-archive-org-7310 13 25 have have VBP web-archive-org-7310 13 26 been be VBN web-archive-org-7310 13 27 performed perform VBN web-archive-org-7310 13 28 to to IN web-archive-org-7310 13 29 date date NN web-archive-org-7310 13 30 to to TO web-archive-org-7310 13 31 answer answer VB web-archive-org-7310 13 32 them -PRON- PRP web-archive-org-7310 13 33 . . . web-archive-org-7310 14 1 One one CD web-archive-org-7310 14 2 notable notable JJ web-archive-org-7310 14 3 exception exception NN web-archive-org-7310 14 4 is be VBZ web-archive-org-7310 14 5 a a DT web-archive-org-7310 14 6 study study NN web-archive-org-7310 14 7 by by IN web-archive-org-7310 14 8 Cho Cho NNP web-archive-org-7310 14 9 and and CC web-archive-org-7310 14 10 Garcia Garcia NNP web-archive-org-7310 14 11 - - HYPH web-archive-org-7310 14 12 Molina Molina NNP web-archive-org-7310 14 13 , , , web-archive-org-7310 14 14 who who WP web-archive-org-7310 14 15 crawled crawl VBD web-archive-org-7310 14 16 a a DT web-archive-org-7310 14 17 set set NN web-archive-org-7310 14 18 of of IN web-archive-org-7310 14 19 720,000 720,000 CD web-archive-org-7310 14 20 pages page NNS web-archive-org-7310 14 21 on on IN web-archive-org-7310 14 22 a a DT web-archive-org-7310 14 23 daily daily JJ web-archive-org-7310 14 24 basis basis NN web-archive-org-7310 14 25 over over IN web-archive-org-7310 14 26 four four CD web-archive-org-7310 14 27 months month NNS web-archive-org-7310 14 28 , , , web-archive-org-7310 14 29 and and CC web-archive-org-7310 14 30 counted count VBN web-archive-org-7310 14 31 pages page NNS web-archive-org-7310 14 32 as as IN web-archive-org-7310 14 33 having have VBG web-archive-org-7310 14 34 changed change VBN web-archive-org-7310 14 35 if if IN web-archive-org-7310 14 36 their -PRON- PRP$ web-archive-org-7310 14 37 MD5 MD5 NNP web-archive-org-7310 14 38 checksum checksum NN web-archive-org-7310 14 39 changed change VBD web-archive-org-7310 14 40 . . . web-archive-org-7310 15 1 They -PRON- PRP web-archive-org-7310 15 2 found find VBD web-archive-org-7310 15 3 that that IN web-archive-org-7310 15 4 40 40 CD web-archive-org-7310 15 5 % % NN web-archive-org-7310 15 6 of of IN web-archive-org-7310 15 7 all all DT web-archive-org-7310 15 8 web web NN web-archive-org-7310 15 9 pages page NNS web-archive-org-7310 15 10 in in IN web-archive-org-7310 15 11 their -PRON- PRP$ web-archive-org-7310 15 12 set set NN web-archive-org-7310 15 13 changed change VBD web-archive-org-7310 15 14 within within IN web-archive-org-7310 15 15 a a DT web-archive-org-7310 15 16 week week NN web-archive-org-7310 15 17 , , , web-archive-org-7310 15 18 and and CC web-archive-org-7310 15 19 23 23 CD web-archive-org-7310 15 20 % % NN web-archive-org-7310 15 21 of of IN web-archive-org-7310 15 22 those those DT web-archive-org-7310 15 23 pages page NNS web-archive-org-7310 15 24 that that WDT web-archive-org-7310 15 25 fell fall VBD web-archive-org-7310 15 26 into into IN web-archive-org-7310 15 27 the the DT web-archive-org-7310 15 28 .com .com CD web-archive-org-7310 15 29 domain domain NN web-archive-org-7310 15 30 changed change VBN web-archive-org-7310 15 31 daily daily RB web-archive-org-7310 15 32 . . . web-archive-org-7310 16 1 This this DT web-archive-org-7310 16 2 paper paper NN web-archive-org-7310 16 3 expands expand VBZ web-archive-org-7310 16 4 on on IN web-archive-org-7310 16 5 Cho Cho NNP web-archive-org-7310 16 6 and and CC web-archive-org-7310 16 7 Garcia Garcia NNP web-archive-org-7310 16 8 - - HYPH web-archive-org-7310 16 9 Molina Molina NNP web-archive-org-7310 16 10 's 's POS web-archive-org-7310 16 11 study study NN web-archive-org-7310 16 12 , , , web-archive-org-7310 16 13 both both CC web-archive-org-7310 16 14 in in IN web-archive-org-7310 16 15 terms term NNS web-archive-org-7310 16 16 of of IN web-archive-org-7310 16 17 coverage coverage NN web-archive-org-7310 16 18 and and CC web-archive-org-7310 16 19 in in IN web-archive-org-7310 16 20 terms term NNS web-archive-org-7310 16 21 of of IN web-archive-org-7310 16 22 sensitivity sensitivity NN web-archive-org-7310 16 23 to to TO web-archive-org-7310 16 24 change change VB web-archive-org-7310 16 25 . . . web-archive-org-7310 17 1 We -PRON- PRP web-archive-org-7310 17 2 crawled crawl VBD web-archive-org-7310 17 3 a a DT web-archive-org-7310 17 4 set set NN web-archive-org-7310 17 5 of of IN web-archive-org-7310 17 6 150,836,209 150,836,209 CD web-archive-org-7310 17 7 HTML html NN web-archive-org-7310 17 8 pages page NNS web-archive-org-7310 17 9 once once RB web-archive-org-7310 17 10 every every DT web-archive-org-7310 17 11 week week NN web-archive-org-7310 17 12 , , , web-archive-org-7310 17 13 over over IN web-archive-org-7310 17 14 a a DT web-archive-org-7310 17 15 span span NN web-archive-org-7310 17 16 of of IN web-archive-org-7310 17 17 11 11 CD web-archive-org-7310 17 18 weeks week NNS web-archive-org-7310 17 19 . . . web-archive-org-7310 18 1 For for IN web-archive-org-7310 18 2 each each DT web-archive-org-7310 18 3 page page NN web-archive-org-7310 18 4 , , , web-archive-org-7310 18 5 we -PRON- PRP web-archive-org-7310 18 6 recorded record VBD web-archive-org-7310 18 7 a a DT web-archive-org-7310 18 8 checksum checksum NN web-archive-org-7310 18 9 of of IN web-archive-org-7310 18 10 the the DT web-archive-org-7310 18 11 page page NN web-archive-org-7310 18 12 , , , web-archive-org-7310 18 13 and and CC web-archive-org-7310 18 14 a a DT web-archive-org-7310 18 15 feature feature NN web-archive-org-7310 18 16 vector vector NN web-archive-org-7310 18 17 of of IN web-archive-org-7310 18 18 the the DT web-archive-org-7310 18 19 words word NNS web-archive-org-7310 18 20 on on IN web-archive-org-7310 18 21 the the DT web-archive-org-7310 18 22 page page NN web-archive-org-7310 18 23 , , , web-archive-org-7310 18 24 plus plus CC web-archive-org-7310 18 25 various various JJ web-archive-org-7310 18 26 other other JJ web-archive-org-7310 18 27 data datum NNS web-archive-org-7310 18 28 such such JJ web-archive-org-7310 18 29 as as IN web-archive-org-7310 18 30 the the DT web-archive-org-7310 18 31 page page NN web-archive-org-7310 18 32 length length NN web-archive-org-7310 18 33 , , , web-archive-org-7310 18 34 the the DT web-archive-org-7310 18 35 HTTP http JJ web-archive-org-7310 18 36 status status NN web-archive-org-7310 18 37 code code NN web-archive-org-7310 18 38 , , , web-archive-org-7310 18 39 etc etc FW web-archive-org-7310 18 40 . . . web-archive-org-7310 19 1 Moreover moreover RB web-archive-org-7310 19 2 , , , web-archive-org-7310 19 3 we -PRON- PRP web-archive-org-7310 19 4 pseudo pseudo NN web-archive-org-7310 19 5 - - VBP web-archive-org-7310 19 6 randomly randomly RB web-archive-org-7310 19 7 selected select VBD web-archive-org-7310 19 8 0.1 0.1 CD web-archive-org-7310 19 9 % % NN web-archive-org-7310 19 10 of of IN web-archive-org-7310 19 11 all all DT web-archive-org-7310 19 12 of of IN web-archive-org-7310 19 13 our -PRON- PRP$ web-archive-org-7310 19 14 URLs url NNS web-archive-org-7310 19 15 , , , web-archive-org-7310 19 16 and and CC web-archive-org-7310 19 17 saved save VBD web-archive-org-7310 19 18 the the DT web-archive-org-7310 19 19 full full JJ web-archive-org-7310 19 20 text text NN web-archive-org-7310 19 21 of of IN web-archive-org-7310 19 22 each each DT web-archive-org-7310 19 23 download download NN web-archive-org-7310 19 24 of of IN web-archive-org-7310 19 25 the the DT web-archive-org-7310 19 26 corresponding correspond VBG web-archive-org-7310 19 27 pages page NNS web-archive-org-7310 19 28 . . . web-archive-org-7310 20 1 After after IN web-archive-org-7310 20 2 completion completion NN web-archive-org-7310 20 3 of of IN web-archive-org-7310 20 4 the the DT web-archive-org-7310 20 5 crawl crawl NN web-archive-org-7310 20 6 , , , web-archive-org-7310 20 7 we -PRON- PRP web-archive-org-7310 20 8 analyzed analyze VBD web-archive-org-7310 20 9 the the DT web-archive-org-7310 20 10 degree degree NN web-archive-org-7310 20 11 of of IN web-archive-org-7310 20 12 change change NN web-archive-org-7310 20 13 of of IN web-archive-org-7310 20 14 each each DT web-archive-org-7310 20 15 page page NN web-archive-org-7310 20 16 , , , web-archive-org-7310 20 17 and and CC web-archive-org-7310 20 18 investigated investigate VBN web-archive-org-7310 20 19 which which WDT web-archive-org-7310 20 20 factors factor NNS web-archive-org-7310 20 21 are be VBP web-archive-org-7310 20 22 correlated correlate VBN web-archive-org-7310 20 23 with with IN web-archive-org-7310 20 24 change change NN web-archive-org-7310 20 25 intensity intensity NN web-archive-org-7310 20 26 . . . web-archive-org-7310 21 1 We -PRON- PRP web-archive-org-7310 21 2 found find VBD web-archive-org-7310 21 3 that that IN web-archive-org-7310 21 4 the the DT web-archive-org-7310 21 5 average average JJ web-archive-org-7310 21 6 degree degree NN web-archive-org-7310 21 7 of of IN web-archive-org-7310 21 8 change change NN web-archive-org-7310 21 9 varies vary VBZ web-archive-org-7310 21 10 widely widely RB web-archive-org-7310 21 11 across across IN web-archive-org-7310 21 12 top top JJ web-archive-org-7310 21 13 - - HYPH web-archive-org-7310 21 14 level level NN web-archive-org-7310 21 15 domains domain NNS web-archive-org-7310 21 16 , , , web-archive-org-7310 21 17 and and CC web-archive-org-7310 21 18 that that IN web-archive-org-7310 21 19 larger large JJR web-archive-org-7310 21 20 pages page NNS web-archive-org-7310 21 21 change change VBP web-archive-org-7310 21 22 more more RBR web-archive-org-7310 21 23 often often RB web-archive-org-7310 21 24 and and CC web-archive-org-7310 21 25 more more RBR web-archive-org-7310 21 26 severely severely RB web-archive-org-7310 21 27 than than IN web-archive-org-7310 21 28 smaller small JJR web-archive-org-7310 21 29 ones one NNS web-archive-org-7310 21 30 . . . web-archive-org-7310 22 1 This this DT web-archive-org-7310 22 2 paper paper NN web-archive-org-7310 22 3 describes describe VBZ web-archive-org-7310 22 4 the the DT web-archive-org-7310 22 5 crawl crawl NN web-archive-org-7310 22 6 and and CC web-archive-org-7310 22 7 the the DT web-archive-org-7310 22 8 data datum NNS web-archive-org-7310 22 9 transformations transformation NNS web-archive-org-7310 22 10 we -PRON- PRP web-archive-org-7310 22 11 performed perform VBD web-archive-org-7310 22 12 on on IN web-archive-org-7310 22 13 the the DT web-archive-org-7310 22 14 logs log NNS web-archive-org-7310 22 15 , , , web-archive-org-7310 22 16 and and CC web-archive-org-7310 22 17 presents present VBZ web-archive-org-7310 22 18 some some DT web-archive-org-7310 22 19 statistical statistical JJ web-archive-org-7310 22 20 observations observation NNS web-archive-org-7310 22 21 on on IN web-archive-org-7310 22 22 the the DT web-archive-org-7310 22 23 degree degree NN web-archive-org-7310 22 24 of of IN web-archive-org-7310 22 25 change change NN web-archive-org-7310 22 26 of of IN web-archive-org-7310 22 27 different different JJ web-archive-org-7310 22 28 classes class NNS web-archive-org-7310 22 29 of of IN web-archive-org-7310 22 30 pages page NNS web-archive-org-7310 22 31 . . . web-archive-org-7310 23 1 Keywords keyword NNS web-archive-org-7310 23 2 Web web NN web-archive-org-7310 23 3 characterization characterization NN web-archive-org-7310 23 4 , , , web-archive-org-7310 23 5 web web NN web-archive-org-7310 23 6 evolution evolution NN web-archive-org-7310 23 7 , , , web-archive-org-7310 23 8 web web NN web-archive-org-7310 23 9 pages page NNS web-archive-org-7310 23 10 , , , web-archive-org-7310 23 11 rate rate NN web-archive-org-7310 23 12 of of IN web-archive-org-7310 23 13 change change NN web-archive-org-7310 23 14 , , , web-archive-org-7310 23 15 degree degree NN web-archive-org-7310 23 16 of of IN web-archive-org-7310 23 17 change change NN web-archive-org-7310 23 18 1 1 CD web-archive-org-7310 23 19 . . . web-archive-org-7310 24 1 INTRODUCTION introduction VB web-archive-org-7310 24 2 The the DT web-archive-org-7310 24 3 searchable searchable JJ web-archive-org-7310 24 4 web web NN web-archive-org-7310 24 5 and and CC web-archive-org-7310 24 6 the the DT web-archive-org-7310 24 7 search search NN web-archive-org-7310 24 8 engines engine NNS web-archive-org-7310 24 9 which which WDT web-archive-org-7310 24 10 survey survey VBP web-archive-org-7310 24 11 it -PRON- PRP web-archive-org-7310 24 12 have have VBP web-archive-org-7310 24 13 become become VBN web-archive-org-7310 24 14 indispensable indispensable JJ web-archive-org-7310 24 15 tools tool NNS web-archive-org-7310 24 16 for for IN web-archive-org-7310 24 17 information information NN web-archive-org-7310 24 18 discovery discovery NN web-archive-org-7310 24 19 . . . web-archive-org-7310 25 1 From from IN web-archive-org-7310 25 2 academic academic JJ web-archive-org-7310 25 3 researchers researcher NNS web-archive-org-7310 25 4 to to IN web-archive-org-7310 25 5 elementary elementary JJ web-archive-org-7310 25 6 - - HYPH web-archive-org-7310 25 7 school school NN web-archive-org-7310 25 8 students student NNS web-archive-org-7310 25 9 , , , web-archive-org-7310 25 10 from from IN web-archive-org-7310 25 11 cancer cancer NN web-archive-org-7310 25 12 patients patient NNS web-archive-org-7310 25 13 to to IN web-archive-org-7310 25 14 pensioners pensioner NNS web-archive-org-7310 25 15 , , , web-archive-org-7310 25 16 from from IN web-archive-org-7310 25 17 local local JJ web-archive-org-7310 25 18 high high JJ web-archive-org-7310 25 19 - - HYPH web-archive-org-7310 25 20 school school NN web-archive-org-7310 25 21 football football NN web-archive-org-7310 25 22 fans fan NNS web-archive-org-7310 25 23 to to IN web-archive-org-7310 25 24 international international JJ web-archive-org-7310 25 25 travelers traveler NNS web-archive-org-7310 25 26 , , , web-archive-org-7310 25 27 the the DT web-archive-org-7310 25 28 indexed index VBN web-archive-org-7310 25 29 content content NN web-archive-org-7310 25 30 of of IN web-archive-org-7310 25 31 the the DT web-archive-org-7310 25 32 web web NN web-archive-org-7310 25 33 is be VBZ web-archive-org-7310 25 34 becoming become VBG web-archive-org-7310 25 35 the the DT web-archive-org-7310 25 36 primary primary JJ web-archive-org-7310 25 37 research research NN web-archive-org-7310 25 38 tool tool NN web-archive-org-7310 25 39 for for IN web-archive-org-7310 25 40 many many JJ web-archive-org-7310 25 41 . . . web-archive-org-7310 26 1 With with IN web-archive-org-7310 26 2 hundreds hundred NNS web-archive-org-7310 26 3 of of IN web-archive-org-7310 26 4 millions million NNS web-archive-org-7310 26 5 of of IN web-archive-org-7310 26 6 people people NNS web-archive-org-7310 26 7 relying rely VBG web-archive-org-7310 26 8 on on IN web-archive-org-7310 26 9 these these DT web-archive-org-7310 26 10 tools tool NNS web-archive-org-7310 26 11 , , , web-archive-org-7310 26 12 one one CD web-archive-org-7310 26 13 is be VBZ web-archive-org-7310 26 14 led lead VBN web-archive-org-7310 26 15 to to TO web-archive-org-7310 26 16 ask ask VB web-archive-org-7310 26 17 if if IN web-archive-org-7310 26 18 the the DT web-archive-org-7310 26 19 tools tool NNS web-archive-org-7310 26 20 provide provide VBP web-archive-org-7310 26 21 useful useful JJ web-archive-org-7310 26 22 , , , web-archive-org-7310 26 23 up up RB web-archive-org-7310 26 24 - - HYPH web-archive-org-7310 26 25 to to IN web-archive-org-7310 26 26 - - HYPH web-archive-org-7310 26 27 date date NN web-archive-org-7310 26 28 results result NNS web-archive-org-7310 26 29 . . . web-archive-org-7310 27 1 Ideally ideally RB web-archive-org-7310 27 2 , , , web-archive-org-7310 27 3 one one PRP web-archive-org-7310 27 4 would would MD web-archive-org-7310 27 5 like like VB web-archive-org-7310 27 6 the the DT web-archive-org-7310 27 7 entire entire JJ web-archive-org-7310 27 8 index index NN web-archive-org-7310 27 9 of of IN web-archive-org-7310 27 10 a a DT web-archive-org-7310 27 11 search search NN web-archive-org-7310 27 12 engine engine NN web-archive-org-7310 27 13 to to TO web-archive-org-7310 27 14 be be VB web-archive-org-7310 27 15 fresh fresh JJ web-archive-org-7310 27 16 , , , web-archive-org-7310 27 17 that that RB web-archive-org-7310 27 18 is is RB web-archive-org-7310 27 19 , , , web-archive-org-7310 27 20 to to TO web-archive-org-7310 27 21 contain contain VB web-archive-org-7310 27 22 the the DT web-archive-org-7310 27 23 most most RBS web-archive-org-7310 27 24 up up RB web-archive-org-7310 27 25 - - HYPH web-archive-org-7310 27 26 to to IN web-archive-org-7310 27 27 - - HYPH web-archive-org-7310 27 28 date date NN web-archive-org-7310 27 29 version version NN web-archive-org-7310 27 30 of of IN web-archive-org-7310 27 31 a a DT web-archive-org-7310 27 32 web web NN web-archive-org-7310 27 33 page page NN web-archive-org-7310 27 34 . . . web-archive-org-7310 28 1 The the DT web-archive-org-7310 28 2 Google Google NNP web-archive-org-7310 28 3 search search NN web-archive-org-7310 28 4 engine engine NN web-archive-org-7310 28 5 attempts attempt NNS web-archive-org-7310 28 6 to to TO web-archive-org-7310 28 7 maintain maintain VB web-archive-org-7310 28 8 a a DT web-archive-org-7310 28 9 fresh fresh JJ web-archive-org-7310 28 10 index index NN web-archive-org-7310 28 11 by by IN web-archive-org-7310 28 12 crawling crawl VBG web-archive-org-7310 28 13 over over RP web-archive-org-7310 28 14 3 3 CD web-archive-org-7310 28 15 billion billion CD web-archive-org-7310 28 16 pages page NNS web-archive-org-7310 28 17 once once RB web-archive-org-7310 28 18 a a DT web-archive-org-7310 28 19 month month NN web-archive-org-7310 28 20   _SP web-archive-org-7310 28 21 [ [ -LRB- web-archive-org-7310 28 22 7 7 CD web-archive-org-7310 28 23 ] ] -RRB- web-archive-org-7310 28 24 , , , web-archive-org-7310 28 25 with with IN web-archive-org-7310 28 26 more more RBR web-archive-org-7310 28 27 frequent frequent JJ web-archive-org-7310 28 28 crawls crawl NNS web-archive-org-7310 28 29 of of IN web-archive-org-7310 28 30 hand hand NN web-archive-org-7310 28 31 - - HYPH web-archive-org-7310 28 32 selected select VBN web-archive-org-7310 28 33 sites site NNS web-archive-org-7310 28 34 that that WDT web-archive-org-7310 28 35 are be VBP web-archive-org-7310 28 36 known know VBN web-archive-org-7310 28 37 to to TO web-archive-org-7310 28 38 change change VB web-archive-org-7310 28 39 more more RBR web-archive-org-7310 28 40 often often RB web-archive-org-7310 28 41 . . . web-archive-org-7310 29 1 In in IN web-archive-org-7310 29 2 addition addition NN web-archive-org-7310 29 3 , , , web-archive-org-7310 29 4 it -PRON- PRP web-archive-org-7310 29 5 offers offer VBZ web-archive-org-7310 29 6 access access NN web-archive-org-7310 29 7 to to IN web-archive-org-7310 29 8 cached cached JJ web-archive-org-7310 29 9 copies copy NNS web-archive-org-7310 29 10 of of IN web-archive-org-7310 29 11 pages page NNS web-archive-org-7310 29 12 , , , web-archive-org-7310 29 13 to to TO web-archive-org-7310 29 14 obviate obviate VB web-archive-org-7310 29 15 problems problem NNS web-archive-org-7310 29 16 arising arise VBG web-archive-org-7310 29 17 some some DT web-archive-org-7310 29 18 of of IN web-archive-org-7310 29 19 the the DT web-archive-org-7310 29 20 crawled crawled JJ web-archive-org-7310 29 21 URLs url NNS web-archive-org-7310 29 22 being be VBG web-archive-org-7310 29 23 out out RB web-archive-org-7310 29 24 - - HYPH web-archive-org-7310 29 25 of of IN web-archive-org-7310 29 26 - - HYPH web-archive-org-7310 29 27 date date NN web-archive-org-7310 29 28 or or CC web-archive-org-7310 29 29 having have VBG web-archive-org-7310 29 30 disappeared disappear VBD web-archive-org-7310 29 31 entirely entirely RB web-archive-org-7310 29 32 . . . web-archive-org-7310 30 1 To to TO web-archive-org-7310 30 2 improve improve VB web-archive-org-7310 30 3 the the DT web-archive-org-7310 30 4 freshness freshness NN web-archive-org-7310 30 5 of of IN web-archive-org-7310 30 6 results result NNS web-archive-org-7310 30 7 returned return VBN web-archive-org-7310 30 8 by by IN web-archive-org-7310 30 9 search search NN web-archive-org-7310 30 10 engines engine NNS web-archive-org-7310 30 11 and and CC web-archive-org-7310 30 12 allow allow VB web-archive-org-7310 30 13 them -PRON- PRP web-archive-org-7310 30 14 to to TO web-archive-org-7310 30 15 spend spend VB web-archive-org-7310 30 16 more more JJR web-archive-org-7310 30 17 of of IN web-archive-org-7310 30 18 their -PRON- PRP$ web-archive-org-7310 30 19 efforts effort NNS web-archive-org-7310 30 20 crawling crawl VBG web-archive-org-7310 30 21 and and CC web-archive-org-7310 30 22 indexing indexing NN web-archive-org-7310 30 23 pages page NNS web-archive-org-7310 30 24 which which WDT web-archive-org-7310 30 25 have have VBP web-archive-org-7310 30 26 changed change VBN web-archive-org-7310 30 27 , , , web-archive-org-7310 30 28 it -PRON- PRP web-archive-org-7310 30 29 is be VBZ web-archive-org-7310 30 30 interesting interesting JJ web-archive-org-7310 30 31 and and CC web-archive-org-7310 30 32 important important JJ web-archive-org-7310 30 33 to to TO web-archive-org-7310 30 34 answer answer VB web-archive-org-7310 30 35 some some DT web-archive-org-7310 30 36 questions question NNS web-archive-org-7310 30 37 about about IN web-archive-org-7310 30 38 the the DT web-archive-org-7310 30 39 dynamic dynamic JJ web-archive-org-7310 30 40 nature nature NN web-archive-org-7310 30 41 of of IN web-archive-org-7310 30 42 the the DT web-archive-org-7310 30 43 web web NN web-archive-org-7310 30 44 . . . web-archive-org-7310 31 1 How how WRB web-archive-org-7310 31 2 fast fast RB web-archive-org-7310 31 3 does do VBZ web-archive-org-7310 31 4 the the DT web-archive-org-7310 31 5 web web NN web-archive-org-7310 31 6 change change VB web-archive-org-7310 31 7 ? ? . web-archive-org-7310 32 1 Does do VBZ web-archive-org-7310 32 2 most most JJS web-archive-org-7310 32 3 of of IN web-archive-org-7310 32 4 the the DT web-archive-org-7310 32 5 content content NN web-archive-org-7310 32 6 remain remain VBP web-archive-org-7310 32 7 unchanged unchanged JJ web-archive-org-7310 32 8 once once IN web-archive-org-7310 32 9 it -PRON- PRP web-archive-org-7310 32 10 has have VBZ web-archive-org-7310 32 11 been be VBN web-archive-org-7310 32 12 authored author VBN web-archive-org-7310 32 13 , , , web-archive-org-7310 32 14 or or CC web-archive-org-7310 32 15 are be VBP web-archive-org-7310 32 16 the the DT web-archive-org-7310 32 17 documents document NNS web-archive-org-7310 32 18 being be VBG web-archive-org-7310 32 19 continuously continuously RB web-archive-org-7310 32 20 updated update VBN web-archive-org-7310 32 21 ? ? . web-archive-org-7310 33 1 Do do VBP web-archive-org-7310 33 2 pages page NNS web-archive-org-7310 33 3 change change VB web-archive-org-7310 33 4 a a DT web-archive-org-7310 33 5 little little JJ web-archive-org-7310 33 6 or or CC web-archive-org-7310 33 7 a a DT web-archive-org-7310 33 8 lot lot NN web-archive-org-7310 33 9 ? ? . web-archive-org-7310 34 1 Is be VBZ web-archive-org-7310 34 2 the the DT web-archive-org-7310 34 3 extent extent NN web-archive-org-7310 34 4 of of IN web-archive-org-7310 34 5 change change NN web-archive-org-7310 34 6 correlated correlate VBD web-archive-org-7310 34 7 to to IN web-archive-org-7310 34 8 any any DT web-archive-org-7310 34 9 other other JJ web-archive-org-7310 34 10 property property NN web-archive-org-7310 34 11 of of IN web-archive-org-7310 34 12 the the DT web-archive-org-7310 34 13 page page NN web-archive-org-7310 34 14 ? ? . web-archive-org-7310 35 1 Do do VBP web-archive-org-7310 35 2 pages page NNS web-archive-org-7310 35 3 change change VB web-archive-org-7310 35 4 and and CC web-archive-org-7310 35 5 then then RB web-archive-org-7310 35 6 change change VB web-archive-org-7310 35 7 back back RP web-archive-org-7310 35 8 ? ? . web-archive-org-7310 36 1 How how WRB web-archive-org-7310 36 2 consistent consistent JJ web-archive-org-7310 36 3 are be VBP web-archive-org-7310 36 4 mirrors mirror NNS web-archive-org-7310 36 5 and and CC web-archive-org-7310 36 6 near near JJ web-archive-org-7310 36 7 - - HYPH web-archive-org-7310 36 8 mirrors mirror NNS web-archive-org-7310 36 9 of of IN web-archive-org-7310 36 10 pages page NNS web-archive-org-7310 36 11 ? ? . web-archive-org-7310 37 1 Questions question NNS web-archive-org-7310 37 2 like like IN web-archive-org-7310 37 3 these these DT web-archive-org-7310 37 4 are be VBP web-archive-org-7310 37 5 of of IN web-archive-org-7310 37 6 great great JJ web-archive-org-7310 37 7 relevance relevance NN web-archive-org-7310 37 8 to to IN web-archive-org-7310 37 9 search search NN web-archive-org-7310 37 10 engines engine NNS web-archive-org-7310 37 11 , , , web-archive-org-7310 37 12 and and CC web-archive-org-7310 37 13 more more RBR web-archive-org-7310 37 14 generally generally RB web-archive-org-7310 37 15 to to IN web-archive-org-7310 37 16 any any DT web-archive-org-7310 37 17 party party NN web-archive-org-7310 37 18 trying try VBG web-archive-org-7310 37 19 to to TO web-archive-org-7310 37 20 maintain maintain VB web-archive-org-7310 37 21 an an DT web-archive-org-7310 37 22 up up NN web-archive-org-7310 37 23 - - HYPH web-archive-org-7310 37 24 to to IN web-archive-org-7310 37 25 - - HYPH web-archive-org-7310 37 26 date date NN web-archive-org-7310 37 27 view view NN web-archive-org-7310 37 28 of of IN web-archive-org-7310 37 29 the the DT web-archive-org-7310 37 30 web web NN web-archive-org-7310 37 31 , , , web-archive-org-7310 37 32 but but CC web-archive-org-7310 37 33 they -PRON- PRP web-archive-org-7310 37 34 are be VBP web-archive-org-7310 37 35 also also RB web-archive-org-7310 37 36 interesting interesting JJ web-archive-org-7310 37 37 in in IN web-archive-org-7310 37 38 their -PRON- PRP$ web-archive-org-7310 37 39 own own JJ web-archive-org-7310 37 40 right right NN web-archive-org-7310 37 41 , , , web-archive-org-7310 37 42 as as IN web-archive-org-7310 37 43 they -PRON- PRP web-archive-org-7310 37 44 shed shed VBD web-archive-org-7310 37 45 light light NN web-archive-org-7310 37 46 on on IN web-archive-org-7310 37 47 the the DT web-archive-org-7310 37 48 evolution evolution NN web-archive-org-7310 37 49 of of IN web-archive-org-7310 37 50 a a DT web-archive-org-7310 37 51 major major JJ web-archive-org-7310 37 52 sociological sociological JJ web-archive-org-7310 37 53 phenomenon phenomenon NN web-archive-org-7310 37 54 : : : web-archive-org-7310 37 55 the the DT web-archive-org-7310 37 56 largest large JJS web-archive-org-7310 37 57 collectively collectively RB web-archive-org-7310 37 58 constructed construct VBN web-archive-org-7310 37 59 information information NN web-archive-org-7310 37 60 repository repository NN web-archive-org-7310 37 61 known know VBN web-archive-org-7310 37 62 to to IN web-archive-org-7310 37 63 man man NN web-archive-org-7310 37 64 . . . web-archive-org-7310 38 1 In in IN web-archive-org-7310 38 2 this this DT web-archive-org-7310 38 3 paper paper NN web-archive-org-7310 38 4 , , , web-archive-org-7310 38 5 we -PRON- PRP web-archive-org-7310 38 6 attempt attempt VBP web-archive-org-7310 38 7 to to TO web-archive-org-7310 38 8 answer answer VB web-archive-org-7310 38 9 some some DT web-archive-org-7310 38 10 of of IN web-archive-org-7310 38 11 these these DT web-archive-org-7310 38 12 questions question NNS web-archive-org-7310 38 13 . . . web-archive-org-7310 39 1 We -PRON- PRP web-archive-org-7310 39 2 recount recount VBP web-archive-org-7310 39 3 how how WRB web-archive-org-7310 39 4 we -PRON- PRP web-archive-org-7310 39 5 collected collect VBD web-archive-org-7310 39 6 151 151 CD web-archive-org-7310 39 7 million million CD web-archive-org-7310 39 8 web web NN web-archive-org-7310 39 9 pages page NNS web-archive-org-7310 39 10 eleven eleven CD web-archive-org-7310 39 11 times time NNS web-archive-org-7310 39 12 over over RB web-archive-org-7310 39 13 , , , web-archive-org-7310 39 14 retaining retain VBG web-archive-org-7310 39 15 salient salient NN web-archive-org-7310 39 16 information information NN web-archive-org-7310 39 17 including include VBG web-archive-org-7310 39 18 a a DT web-archive-org-7310 39 19 feature feature NN web-archive-org-7310 39 20 vector vector NN web-archive-org-7310 39 21 of of IN web-archive-org-7310 39 22 each each DT web-archive-org-7310 39 23 page page NN web-archive-org-7310 39 24 . . . web-archive-org-7310 40 1 We -PRON- PRP web-archive-org-7310 40 2 describe describe VBP web-archive-org-7310 40 3 how how WRB web-archive-org-7310 40 4 we -PRON- PRP web-archive-org-7310 40 5 distilled distil VBD web-archive-org-7310 40 6 the the DT web-archive-org-7310 40 7 collected collected JJ web-archive-org-7310 40 8 information information NN web-archive-org-7310 40 9 about about IN web-archive-org-7310 40 10 each each DT web-archive-org-7310 40 11 URL url NN web-archive-org-7310 40 12 into into IN web-archive-org-7310 40 13 a a DT web-archive-org-7310 40 14 summary summary NN web-archive-org-7310 40 15 record record NN web-archive-org-7310 40 16 , , , web-archive-org-7310 40 17 tabulating tabulate VBG web-archive-org-7310 40 18 the the DT web-archive-org-7310 40 19 feature feature NN web-archive-org-7310 40 20 vectors vector NNS web-archive-org-7310 40 21 . . . web-archive-org-7310 41 1 We -PRON- PRP web-archive-org-7310 41 2 sketch sketch VBP web-archive-org-7310 41 3 the the DT web-archive-org-7310 41 4 framework framework NN web-archive-org-7310 41 5 we -PRON- PRP web-archive-org-7310 41 6 used use VBD web-archive-org-7310 41 7 to to TO web-archive-org-7310 41 8 mine -PRON- PRP web-archive-org-7310 41 9 the the DT web-archive-org-7310 41 10 distilled distil VBN web-archive-org-7310 41 11 data datum NNS web-archive-org-7310 41 12 for for IN web-archive-org-7310 41 13 statistical statistical JJ web-archive-org-7310 41 14 information information NN web-archive-org-7310 41 15 . . . web-archive-org-7310 42 1 We -PRON- PRP web-archive-org-7310 42 2 present present VBP web-archive-org-7310 42 3 the the DT web-archive-org-7310 42 4 most most RBS web-archive-org-7310 42 5 interesting interesting JJ web-archive-org-7310 42 6 results result NNS web-archive-org-7310 42 7 of of IN web-archive-org-7310 42 8 this this DT web-archive-org-7310 42 9 data datum NNS web-archive-org-7310 42 10 mining mining NN web-archive-org-7310 42 11 . . . web-archive-org-7310 43 1 Finally finally RB web-archive-org-7310 43 2 , , , web-archive-org-7310 43 3 we -PRON- PRP web-archive-org-7310 43 4 draw draw VBP web-archive-org-7310 43 5 some some DT web-archive-org-7310 43 6 conclusions conclusion NNS web-archive-org-7310 43 7 and and CC web-archive-org-7310 43 8 offer offer VBP web-archive-org-7310 43 9 avenues avenue NNS web-archive-org-7310 43 10 of of IN web-archive-org-7310 43 11 future future JJ web-archive-org-7310 43 12 work work NN web-archive-org-7310 43 13 . . . web-archive-org-7310 44 1 2 2 LS web-archive-org-7310 44 2 . . . web-archive-org-7310 45 1 RELATED related JJ web-archive-org-7310 45 2 WORK work NN web-archive-org-7310 45 3 This this DT web-archive-org-7310 45 4 paper paper NN web-archive-org-7310 45 5 expands expand VBZ web-archive-org-7310 45 6 on on IN web-archive-org-7310 45 7 a a DT web-archive-org-7310 45 8 study study NN web-archive-org-7310 45 9 by by IN web-archive-org-7310 45 10 Cho Cho NNP web-archive-org-7310 45 11 and and CC web-archive-org-7310 45 12 Garcia Garcia NNP web-archive-org-7310 45 13 - - HYPH web-archive-org-7310 45 14 Molina Molina NNP web-archive-org-7310 45 15   _SP web-archive-org-7310 45 16 [ [ -LRB- web-archive-org-7310 45 17 5 5 CD web-archive-org-7310 45 18 ] ] -RRB- web-archive-org-7310 45 19 . . . web-archive-org-7310 46 1 The the DT web-archive-org-7310 46 2 authors author NNS web-archive-org-7310 46 3 of of IN web-archive-org-7310 46 4 that that DT web-archive-org-7310 46 5 study study NN web-archive-org-7310 46 6 downloaded download VBD web-archive-org-7310 46 7 720,000 720,000 CD web-archive-org-7310 46 8 pages page NNS web-archive-org-7310 46 9 drawn draw VBN web-archive-org-7310 46 10 from from IN web-archive-org-7310 46 11 270 270 CD web-archive-org-7310 46 12 " " `` web-archive-org-7310 46 13 popular popular JJ web-archive-org-7310 46 14 " " '' web-archive-org-7310 46 15 web web NN web-archive-org-7310 46 16 servers server NNS web-archive-org-7310 46 17 ( ( -LRB- web-archive-org-7310 46 18 not not RB web-archive-org-7310 46 19 exceeding exceed VBG web-archive-org-7310 46 20 3,000 3,000 CD web-archive-org-7310 46 21 pages page NNS web-archive-org-7310 46 22 per per IN web-archive-org-7310 46 23 server server NN web-archive-org-7310 46 24 ) ) -RRB- web-archive-org-7310 46 25 on on IN web-archive-org-7310 46 26 a a DT web-archive-org-7310 46 27 daily daily JJ web-archive-org-7310 46 28 basis basis NN web-archive-org-7310 46 29 over over IN web-archive-org-7310 46 30 the the DT web-archive-org-7310 46 31 course course NN web-archive-org-7310 46 32 of of IN web-archive-org-7310 46 33 four four CD web-archive-org-7310 46 34 months month NNS web-archive-org-7310 46 35 , , , web-archive-org-7310 46 36 and and CC web-archive-org-7310 46 37 retained retain VBD web-archive-org-7310 46 38 the the DT web-archive-org-7310 46 39 MD5 MD5 NNP web-archive-org-7310 46 40 checksum checksum NN web-archive-org-7310 46 41 of of IN web-archive-org-7310 46 42 the the DT web-archive-org-7310 46 43 contents content NNS web-archive-org-7310 46 44 ( ( -LRB- web-archive-org-7310 46 45 including include VBG web-archive-org-7310 46 46 the the DT web-archive-org-7310 46 47 HTML HTML NNP web-archive-org-7310 46 48 markup markup NN web-archive-org-7310 46 49 ) ) -RRB- web-archive-org-7310 46 50 of of IN web-archive-org-7310 46 51 each each DT web-archive-org-7310 46 52 page page NN web-archive-org-7310 46 53 . . . web-archive-org-7310 47 1 This this DT web-archive-org-7310 47 2 allowed allow VBD web-archive-org-7310 47 3 them -PRON- PRP web-archive-org-7310 47 4 to to TO web-archive-org-7310 47 5 determine determine VB web-archive-org-7310 47 6 if if IN web-archive-org-7310 47 7 a a DT web-archive-org-7310 47 8 document document NN web-archive-org-7310 47 9 had have VBD web-archive-org-7310 47 10 changed change VBN web-archive-org-7310 47 11 , , , web-archive-org-7310 47 12 although although IN web-archive-org-7310 47 13 it -PRON- PRP web-archive-org-7310 47 14 did do VBD web-archive-org-7310 47 15 not not RB web-archive-org-7310 47 16 allow allow VB web-archive-org-7310 47 17 them -PRON- PRP web-archive-org-7310 47 18 to to TO web-archive-org-7310 47 19 assess assess VB web-archive-org-7310 47 20 the the DT web-archive-org-7310 47 21 degree degree NN web-archive-org-7310 47 22 of of IN web-archive-org-7310 47 23 change change NN web-archive-org-7310 47 24 . . . web-archive-org-7310 48 1 Among among IN web-archive-org-7310 48 2 other other JJ web-archive-org-7310 48 3 things thing NNS web-archive-org-7310 48 4 , , , web-archive-org-7310 48 5 they -PRON- PRP web-archive-org-7310 48 6 found find VBD web-archive-org-7310 48 7 that that IN web-archive-org-7310 48 8 pages page NNS web-archive-org-7310 48 9 drawn draw VBN web-archive-org-7310 48 10 from from IN web-archive-org-7310 48 11 servers server NNS web-archive-org-7310 48 12 in in IN web-archive-org-7310 48 13 the the DT web-archive-org-7310 48 14 .com .com CD web-archive-org-7310 48 15 domain domain NN web-archive-org-7310 48 16 changed change VBD web-archive-org-7310 48 17 substantially substantially RB web-archive-org-7310 48 18 faster fast RBR web-archive-org-7310 48 19 than than IN web-archive-org-7310 48 20 those those DT web-archive-org-7310 48 21 in in IN web-archive-org-7310 48 22 other other JJ web-archive-org-7310 48 23 domains domain NNS web-archive-org-7310 48 24 , , , web-archive-org-7310 48 25 while while IN web-archive-org-7310 48 26 pages page NNS web-archive-org-7310 48 27 in in IN web-archive-org-7310 48 28 the the DT web-archive-org-7310 48 29 .gov .gov CD web-archive-org-7310 48 30 domain domain NN web-archive-org-7310 48 31 changed change VBD web-archive-org-7310 48 32 substantially substantially RB web-archive-org-7310 48 33 slower slow RBR web-archive-org-7310 48 34 . . . web-archive-org-7310 49 1 Overall overall RB web-archive-org-7310 49 2 , , , web-archive-org-7310 49 3 they -PRON- PRP web-archive-org-7310 49 4 found find VBD web-archive-org-7310 49 5 that that IN web-archive-org-7310 49 6 about about RB web-archive-org-7310 49 7 40 40 CD web-archive-org-7310 49 8 % % NN web-archive-org-7310 49 9 of of IN web-archive-org-7310 49 10 all all DT web-archive-org-7310 49 11 web web NN web-archive-org-7310 49 12 pages page NNS web-archive-org-7310 49 13 changed change VBN web-archive-org-7310 49 14 within within IN web-archive-org-7310 49 15 a a DT web-archive-org-7310 49 16 week week NN web-archive-org-7310 49 17 , , , web-archive-org-7310 49 18 and and CC web-archive-org-7310 49 19 that that IN web-archive-org-7310 49 20 it -PRON- PRP web-archive-org-7310 49 21 took take VBD web-archive-org-7310 49 22 about about RB web-archive-org-7310 49 23 50 50 CD web-archive-org-7310 49 24 days day NNS web-archive-org-7310 49 25 for for IN web-archive-org-7310 49 26 half half NN web-archive-org-7310 49 27 of of IN web-archive-org-7310 49 28 all all DT web-archive-org-7310 49 29 pages page NNS web-archive-org-7310 49 30 to to TO web-archive-org-7310 49 31 have have VB web-archive-org-7310 49 32 changed change VBN web-archive-org-7310 49 33 . . . web-archive-org-7310 50 1 They -PRON- PRP web-archive-org-7310 50 2 also also RB web-archive-org-7310 50 3 found find VBD web-archive-org-7310 50 4 that that IN web-archive-org-7310 50 5 almost almost RB web-archive-org-7310 50 6 25 25 CD web-archive-org-7310 50 7 % % NN web-archive-org-7310 50 8 of of IN web-archive-org-7310 50 9 the the DT web-archive-org-7310 50 10 pages page NNS web-archive-org-7310 50 11 in in IN web-archive-org-7310 50 12 .com .com , web-archive-org-7310 50 13 changed change VBN web-archive-org-7310 50 14 within within IN web-archive-org-7310 50 15 a a DT web-archive-org-7310 50 16 day day NN web-archive-org-7310 50 17 , , , web-archive-org-7310 50 18 and and CC web-archive-org-7310 50 19 that that IN web-archive-org-7310 50 20 it -PRON- PRP web-archive-org-7310 50 21 took take VBD web-archive-org-7310 50 22 11 11 CD web-archive-org-7310 50 23 days day NNS web-archive-org-7310 50 24 for for IN web-archive-org-7310 50 25 half half NN web-archive-org-7310 50 26 of of IN web-archive-org-7310 50 27 all all DT web-archive-org-7310 50 28 .com .com NFP web-archive-org-7310 50 29 pages page NNS web-archive-org-7310 50 30 to to TO web-archive-org-7310 50 31 have have VB web-archive-org-7310 50 32 changed change VBN web-archive-org-7310 50 33 . . . web-archive-org-7310 51 1 By by IN web-archive-org-7310 51 2 contrast contrast NN web-archive-org-7310 51 3 , , , web-archive-org-7310 51 4 it -PRON- PRP web-archive-org-7310 51 5 took take VBD web-archive-org-7310 51 6 four four CD web-archive-org-7310 51 7 months month NNS web-archive-org-7310 51 8 ( ( -LRB- web-archive-org-7310 51 9 the the DT web-archive-org-7310 51 10 duration duration NN web-archive-org-7310 51 11 of of IN web-archive-org-7310 51 12 their -PRON- PRP$ web-archive-org-7310 51 13 study study NN web-archive-org-7310 51 14 ) ) -RRB- web-archive-org-7310 51 15 for for IN web-archive-org-7310 51 16 half half NN web-archive-org-7310 51 17 of of IN web-archive-org-7310 51 18 the the DT web-archive-org-7310 51 19 .gov .gov CD web-archive-org-7310 51 20 pages page NNS web-archive-org-7310 51 21 to to TO web-archive-org-7310 51 22 have have VB web-archive-org-7310 51 23 changed change VBN web-archive-org-7310 51 24 . . . web-archive-org-7310 52 1 Sun Sun NNP web-archive-org-7310 52 2 et et NNP web-archive-org-7310 52 3 al al NNP web-archive-org-7310 52 4 . . . web-archive-org-7310 53 1 [ [ -LRB- web-archive-org-7310 53 2 11 11 CD web-archive-org-7310 53 3 ] ] -RRB- web-archive-org-7310 53 4 studied study VBD web-archive-org-7310 53 5 the the DT web-archive-org-7310 53 6 efficacy efficacy NN web-archive-org-7310 53 7 of of IN web-archive-org-7310 53 8 web web NN web-archive-org-7310 53 9 anonymizers anonymizer NNS web-archive-org-7310 53 10 . . . web-archive-org-7310 54 1 As as IN web-archive-org-7310 54 2 part part NN web-archive-org-7310 54 3 of of IN web-archive-org-7310 54 4 that that DT web-archive-org-7310 54 5 study study NN web-archive-org-7310 54 6 , , , web-archive-org-7310 54 7 they -PRON- PRP web-archive-org-7310 54 8 drew draw VBD web-archive-org-7310 54 9 a a DT web-archive-org-7310 54 10 set set NN web-archive-org-7310 54 11 of of IN web-archive-org-7310 54 12 100,000 100,000 CD web-archive-org-7310 54 13 web web NN web-archive-org-7310 54 14 pages page NNS web-archive-org-7310 54 15 from from IN web-archive-org-7310 54 16 the the DT web-archive-org-7310 54 17 Open Open NNP web-archive-org-7310 54 18 Directory Directory NNP web-archive-org-7310 54 19 listing listing NN web-archive-org-7310 54 20 , , , web-archive-org-7310 54 21 and and CC web-archive-org-7310 54 22 crawled crawl VBD web-archive-org-7310 54 23 each each DT web-archive-org-7310 54 24 page page NN web-archive-org-7310 54 25 twice twice RB web-archive-org-7310 54 26 , , , web-archive-org-7310 54 27 with with IN web-archive-org-7310 54 28 the the DT web-archive-org-7310 54 29 second second JJ web-archive-org-7310 54 30 retrieval retrieval NN web-archive-org-7310 54 31 immediately immediately RB web-archive-org-7310 54 32 following follow VBG web-archive-org-7310 54 33 the the DT web-archive-org-7310 54 34 first first JJ web-archive-org-7310 54 35 one one CD web-archive-org-7310 54 36 . . . web-archive-org-7310 55 1 Since since IN web-archive-org-7310 55 2 they -PRON- PRP web-archive-org-7310 55 3 were be VBD web-archive-org-7310 55 4 interested interested JJ web-archive-org-7310 55 5 in in IN web-archive-org-7310 55 6 the the DT web-archive-org-7310 55 7 information information NN web-archive-org-7310 55 8 leakage leakage NN web-archive-org-7310 55 9 of of IN web-archive-org-7310 55 10 encrypted encrypt VBN web-archive-org-7310 55 11 channels channel NNS web-archive-org-7310 55 12 , , , web-archive-org-7310 55 13 they -PRON- PRP web-archive-org-7310 55 14 did do VBD web-archive-org-7310 55 15 not not RB web-archive-org-7310 55 16 compare compare VB web-archive-org-7310 55 17 checksums checksum NNS web-archive-org-7310 55 18 of of IN web-archive-org-7310 55 19 the the DT web-archive-org-7310 55 20 returned return VBN web-archive-org-7310 55 21 pages page NNS web-archive-org-7310 55 22 ; ; : web-archive-org-7310 55 23 rather rather RB web-archive-org-7310 55 24 , , , web-archive-org-7310 55 25 they -PRON- PRP web-archive-org-7310 55 26 compared compare VBD web-archive-org-7310 55 27 the the DT web-archive-org-7310 55 28 lengths length NNS web-archive-org-7310 55 29 of of IN web-archive-org-7310 55 30 the the DT web-archive-org-7310 55 31 pages page NNS web-archive-org-7310 55 32 and and CC web-archive-org-7310 55 33 the the DT web-archive-org-7310 55 34 number number NN web-archive-org-7310 55 35 and and CC web-archive-org-7310 55 36 lengths length NNS web-archive-org-7310 55 37 of of IN web-archive-org-7310 55 38 their -PRON- PRP$ web-archive-org-7310 55 39 embedded embed VBN web-archive-org-7310 55 40 images image NNS web-archive-org-7310 55 41 and and CC web-archive-org-7310 55 42 frames frame NNS web-archive-org-7310 55 43 ( ( -LRB- web-archive-org-7310 55 44 which which WDT web-archive-org-7310 55 45 will will MD web-archive-org-7310 55 46 appear appear VB web-archive-org-7310 55 47 to to IN web-archive-org-7310 55 48 an an DT web-archive-org-7310 55 49 eavesdropper eavesdropper NN web-archive-org-7310 55 50 as as IN web-archive-org-7310 55 51 temporally temporally RB web-archive-org-7310 55 52 closely closely RB web-archive-org-7310 55 53 spaced space VBN web-archive-org-7310 55 54 TCP TCP NNP web-archive-org-7310 55 55 packets packet NNS web-archive-org-7310 55 56 ) ) -RRB- web-archive-org-7310 55 57 . . . web-archive-org-7310 56 1 They -PRON- PRP web-archive-org-7310 56 2 found find VBD web-archive-org-7310 56 3 that that IN web-archive-org-7310 56 4 40 40 CD web-archive-org-7310 56 5 % % NN web-archive-org-7310 56 6 of of IN web-archive-org-7310 56 7 pages page NNS web-archive-org-7310 56 8 changed change VBD web-archive-org-7310 56 9 signatures signature NNS web-archive-org-7310 56 10 , , , web-archive-org-7310 56 11 and and CC web-archive-org-7310 56 12 that that IN web-archive-org-7310 56 13 14 14 CD web-archive-org-7310 56 14 % % NN web-archive-org-7310 56 15 of of IN web-archive-org-7310 56 16 pages page NNS web-archive-org-7310 56 17 changed change VBN web-archive-org-7310 56 18 by by IN web-archive-org-7310 56 19 30 30 CD web-archive-org-7310 56 20 % % NN web-archive-org-7310 56 21 or or CC web-archive-org-7310 56 22 more more JJR web-archive-org-7310 56 23 , , , web-archive-org-7310 56 24 using use VBG web-archive-org-7310 56 25 a a DT web-archive-org-7310 56 26 Jaccard Jaccard NNP web-archive-org-7310 56 27 - - HYPH web-archive-org-7310 56 28 coefficient coefficient NN web-archive-org-7310 56 29 - - HYPH web-archive-org-7310 56 30 based base VBN web-archive-org-7310 56 31 similarity similarity NN web-archive-org-7310 56 32 metric metric NN web-archive-org-7310 56 33 akin akin NNP web-archive-org-7310 56 34 to to IN web-archive-org-7310 56 35 the the DT web-archive-org-7310 56 36 one one NN web-archive-org-7310 56 37 we -PRON- PRP web-archive-org-7310 56 38 use use VBP web-archive-org-7310 56 39 , , , web-archive-org-7310 56 40 but but CC web-archive-org-7310 56 41 based base VBN web-archive-org-7310 56 42 on on IN web-archive-org-7310 56 43 the the DT web-archive-org-7310 56 44 size size NN web-archive-org-7310 56 45 of of IN web-archive-org-7310 56 46 a a DT web-archive-org-7310 56 47 document document NN web-archive-org-7310 56 48 including include VBG web-archive-org-7310 56 49 sizes size NNS web-archive-org-7310 56 50 of of IN web-archive-org-7310 56 51 embedded embed VBN web-archive-org-7310 56 52 images image NNS web-archive-org-7310 56 53 , , , web-archive-org-7310 56 54 not not RB web-archive-org-7310 56 55 contents content NNS web-archive-org-7310 56 56 . . . web-archive-org-7310 57 1 Douglis Douglis NNP web-archive-org-7310 57 2 et et FW web-archive-org-7310 57 3 al al NNP web-archive-org-7310 57 4 . . . web-archive-org-7310 58 1 [ [ -LRB- web-archive-org-7310 58 2 6 6 CD web-archive-org-7310 58 3 ] ] -RRB- web-archive-org-7310 58 4 studied study VBD web-archive-org-7310 58 5 a a DT web-archive-org-7310 58 6 web web NN web-archive-org-7310 58 7 trace trace NN web-archive-org-7310 58 8 consisting consist VBG web-archive-org-7310 58 9 of of IN web-archive-org-7310 58 10 950,000 950,000 CD web-archive-org-7310 58 11 entries entry NNS web-archive-org-7310 58 12 ( ( -LRB- web-archive-org-7310 58 13 each each DT web-archive-org-7310 58 14 entry entry NN web-archive-org-7310 58 15 representing represent VBG web-archive-org-7310 58 16 a a DT web-archive-org-7310 58 17 web web NN web-archive-org-7310 58 18 access access NN web-archive-org-7310 58 19 ) ) -RRB- web-archive-org-7310 58 20 collected collect VBD web-archive-org-7310 58 21 over over IN web-archive-org-7310 58 22 17 17 CD web-archive-org-7310 58 23 days day NNS web-archive-org-7310 58 24 at at IN web-archive-org-7310 58 25 the the DT web-archive-org-7310 58 26 gateway gateway NN web-archive-org-7310 58 27 between between IN web-archive-org-7310 58 28 AT&T AT&T NNP web-archive-org-7310 58 29 Labs Labs NNP web-archive-org-7310 58 30 - - HYPH web-archive-org-7310 58 31 Research Research NNP web-archive-org-7310 58 32 and and CC web-archive-org-7310 58 33 the the DT web-archive-org-7310 58 34 Internet internet NN web-archive-org-7310 58 35 . . . web-archive-org-7310 59 1 They -PRON- PRP web-archive-org-7310 59 2 recorded record VBD web-archive-org-7310 59 3 the the DT web-archive-org-7310 59 4 " " `` web-archive-org-7310 59 5 last last RB web-archive-org-7310 59 6 - - HYPH web-archive-org-7310 59 7 modified modify VBN web-archive-org-7310 59 8 " " '' web-archive-org-7310 59 9 timestamp timestamp NN web-archive-org-7310 59 10 transmitted transmit VBN web-archive-org-7310 59 11 by by IN web-archive-org-7310 59 12 the the DT web-archive-org-7310 59 13 web web NN web-archive-org-7310 59 14 server server NN web-archive-org-7310 59 15 ( ( -LRB- web-archive-org-7310 59 16 implicitly implicitly RB web-archive-org-7310 59 17 assuming assume VBG web-archive-org-7310 59 18 that that IN web-archive-org-7310 59 19 web web NN web-archive-org-7310 59 20 servers server NNS web-archive-org-7310 59 21 transmit transmit VB web-archive-org-7310 59 22 accurate accurate JJ web-archive-org-7310 59 23 information information NN web-archive-org-7310 59 24 ) ) -RRB- web-archive-org-7310 59 25 . . . web-archive-org-7310 60 1 In in IN web-archive-org-7310 60 2 addition addition NN web-archive-org-7310 60 3 , , , web-archive-org-7310 60 4 they -PRON- PRP web-archive-org-7310 60 5 mined mine VBD web-archive-org-7310 60 6 each each DT web-archive-org-7310 60 7 page page NN web-archive-org-7310 60 8 for for IN web-archive-org-7310 60 9 items item NNS web-archive-org-7310 60 10 such such JJ web-archive-org-7310 60 11 as as IN web-archive-org-7310 60 12 phone phone NN web-archive-org-7310 60 13 numbers number NNS web-archive-org-7310 60 14 using use VBG web-archive-org-7310 60 15 a a DT web-archive-org-7310 60 16 domain domain NN web-archive-org-7310 60 17 - - HYPH web-archive-org-7310 60 18 specific specific JJ web-archive-org-7310 60 19 semantic semantic JJ web-archive-org-7310 60 20 analysis analysis NN web-archive-org-7310 60 21 technique technique NN web-archive-org-7310 60 22 called call VBN web-archive-org-7310 60 23 " " `` web-archive-org-7310 60 24 grinking grinke VBG web-archive-org-7310 60 25 " " '' web-archive-org-7310 60 26 , , , web-archive-org-7310 60 27 and and CC web-archive-org-7310 60 28 measured measure VBD web-archive-org-7310 60 29 the the DT web-archive-org-7310 60 30 rate rate NN web-archive-org-7310 60 31 of of IN web-archive-org-7310 60 32 change change NN web-archive-org-7310 60 33 of of IN web-archive-org-7310 60 34 these these DT web-archive-org-7310 60 35 items item NNS web-archive-org-7310 60 36 . . . web-archive-org-7310 61 1 They -PRON- PRP web-archive-org-7310 61 2 found find VBD web-archive-org-7310 61 3 that that IN web-archive-org-7310 61 4 according accord VBG web-archive-org-7310 61 5 to to IN web-archive-org-7310 61 6 the the DT web-archive-org-7310 61 7 last last RB web-archive-org-7310 61 8 - - HYPH web-archive-org-7310 61 9 modified modify VBN web-archive-org-7310 61 10 metric metric JJ web-archive-org-7310 61 11 , , , web-archive-org-7310 61 12 16.5 16.5 CD web-archive-org-7310 61 13 % % NN web-archive-org-7310 61 14 of of IN web-archive-org-7310 61 15 the the DT web-archive-org-7310 61 16 resources resource NNS web-archive-org-7310 61 17 ( ( -LRB- web-archive-org-7310 61 18 including include VBG web-archive-org-7310 61 19 HTML html NN web-archive-org-7310 61 20 pages page NNS web-archive-org-7310 61 21 as as RB web-archive-org-7310 61 22 well well RB web-archive-org-7310 61 23 as as IN web-archive-org-7310 61 24 other other JJ web-archive-org-7310 61 25 content content NN web-archive-org-7310 61 26 , , , web-archive-org-7310 61 27 such such JJ web-archive-org-7310 61 28 as as IN web-archive-org-7310 61 29 images image NNS web-archive-org-7310 61 30 ) ) -RRB- web-archive-org-7310 61 31 that that WDT web-archive-org-7310 61 32 were be VBD web-archive-org-7310 61 33 accessed access VBN web-archive-org-7310 61 34 multiple multiple JJ web-archive-org-7310 61 35 times time NNS web-archive-org-7310 61 36 changed change VBD web-archive-org-7310 61 37 every every DT web-archive-org-7310 61 38 time time NN web-archive-org-7310 61 39 they -PRON- PRP web-archive-org-7310 61 40 were be VBD web-archive-org-7310 61 41 accessed access VBN web-archive-org-7310 61 42 . . . web-archive-org-7310 62 1 They -PRON- PRP web-archive-org-7310 62 2 also also RB web-archive-org-7310 62 3 found find VBD web-archive-org-7310 62 4 that that IN web-archive-org-7310 62 5 among among IN web-archive-org-7310 62 6 the the DT web-archive-org-7310 62 7 HTML html NN web-archive-org-7310 62 8 pages page NNS web-archive-org-7310 62 9 that that WDT web-archive-org-7310 62 10 were be VBD web-archive-org-7310 62 11 accessed access VBN web-archive-org-7310 62 12 more more RBR web-archive-org-7310 62 13 than than IN web-archive-org-7310 62 14 once once RB web-archive-org-7310 62 15 , , , web-archive-org-7310 62 16 almost almost RB web-archive-org-7310 62 17 60 60 CD web-archive-org-7310 62 18 % % NN web-archive-org-7310 62 19 experienced experience VBD web-archive-org-7310 62 20 a a DT web-archive-org-7310 62 21 change change NN web-archive-org-7310 62 22 in in IN web-archive-org-7310 62 23 HREF HREF NNP web-archive-org-7310 62 24 links link NNS web-archive-org-7310 62 25 , , , web-archive-org-7310 62 26 and and CC web-archive-org-7310 62 27 over over IN web-archive-org-7310 62 28 50 50 CD web-archive-org-7310 62 29 % % NN web-archive-org-7310 62 30 of of IN web-archive-org-7310 62 31 them -PRON- PRP web-archive-org-7310 62 32 experienced experience VBD web-archive-org-7310 62 33 a a DT web-archive-org-7310 62 34 change change NN web-archive-org-7310 62 35 in in IN web-archive-org-7310 62 36 IMG IMG NNP web-archive-org-7310 62 37 links link NNS web-archive-org-7310 62 38 . . . web-archive-org-7310 63 1 Brewington Brewington NNP web-archive-org-7310 63 2 and and CC web-archive-org-7310 63 3 Cybenko Cybenko NNP web-archive-org-7310 63 4 built build VBD web-archive-org-7310 63 5 a a DT web-archive-org-7310 63 6 web web NN web-archive-org-7310 63 7 clipping clip VBG web-archive-org-7310 63 8 service service NN web-archive-org-7310 63 9 , , , web-archive-org-7310 63 10 which which WDT web-archive-org-7310 63 11 they -PRON- PRP web-archive-org-7310 63 12 leveraged leverage VBD web-archive-org-7310 63 13 to to TO web-archive-org-7310 63 14 study study VB web-archive-org-7310 63 15 the the DT web-archive-org-7310 63 16 change change NN web-archive-org-7310 63 17 rate rate NN web-archive-org-7310 63 18 of of IN web-archive-org-7310 63 19 web web NN web-archive-org-7310 63 20 pages page NNS web-archive-org-7310 63 21   _SP web-archive-org-7310 63 22 [ [ -LRB- web-archive-org-7310 63 23 1 1 CD web-archive-org-7310 63 24 ] ] -RRB- web-archive-org-7310 63 25 . . . web-archive-org-7310 64 1 They -PRON- PRP web-archive-org-7310 64 2 did do VBD web-archive-org-7310 64 3 so so RB web-archive-org-7310 64 4 by by IN web-archive-org-7310 64 5 recording record VBG web-archive-org-7310 64 6 the the DT web-archive-org-7310 64 7 last last RB web-archive-org-7310 64 8 - - HYPH web-archive-org-7310 64 9 modified modify VBN web-archive-org-7310 64 10 time time NN web-archive-org-7310 64 11 stamp stamp NN web-archive-org-7310 64 12 , , , web-archive-org-7310 64 13 the the DT web-archive-org-7310 64 14 time time NN web-archive-org-7310 64 15 of of IN web-archive-org-7310 64 16 download download NN web-archive-org-7310 64 17 , , , web-archive-org-7310 64 18 and and CC web-archive-org-7310 64 19 various various JJ web-archive-org-7310 64 20 stylistic stylistic JJ web-archive-org-7310 64 21 attributes attribute NNS web-archive-org-7310 64 22 ( ( -LRB- web-archive-org-7310 64 23 number number NN web-archive-org-7310 64 24 of of IN web-archive-org-7310 64 25 images image NNS web-archive-org-7310 64 26 , , , web-archive-org-7310 64 27 links link NNS web-archive-org-7310 64 28 , , , web-archive-org-7310 64 29 tables table NNS web-archive-org-7310 64 30 , , , web-archive-org-7310 64 31 etc etc FW web-archive-org-7310 64 32 ) ) -RRB- web-archive-org-7310 64 33 of of IN web-archive-org-7310 64 34 each each DT web-archive-org-7310 64 35 downloaded download VBN web-archive-org-7310 64 36 HTML html NN web-archive-org-7310 64 37 page page NN web-archive-org-7310 64 38 . . . web-archive-org-7310 65 1 Their -PRON- PRP$ web-archive-org-7310 65 2 service service NN web-archive-org-7310 65 3 downloaded download VBD web-archive-org-7310 65 4 about about IN web-archive-org-7310 65 5 100,000 100,000 CD web-archive-org-7310 65 6 pages page NNS web-archive-org-7310 65 7 per per IN web-archive-org-7310 65 8 day day NN web-archive-org-7310 65 9 , , , web-archive-org-7310 65 10 selected select VBN web-archive-org-7310 65 11 based base VBN web-archive-org-7310 65 12 on on IN web-archive-org-7310 65 13 their -PRON- PRP$ web-archive-org-7310 65 14 topical topical JJ web-archive-org-7310 65 15 interest interest NN web-archive-org-7310 65 16 , , , web-archive-org-7310 65 17 recrawling recrawle VBG web-archive-org-7310 65 18 no no DT web-archive-org-7310 65 19 page page NN web-archive-org-7310 65 20 more more RBR web-archive-org-7310 65 21 often often RB web-archive-org-7310 65 22 than than IN web-archive-org-7310 65 23 once once RB web-archive-org-7310 65 24 every every DT web-archive-org-7310 65 25 three three CD web-archive-org-7310 65 26 days day NNS web-archive-org-7310 65 27 . . . web-archive-org-7310 66 1 They -PRON- PRP web-archive-org-7310 66 2 evaluated evaluate VBD web-archive-org-7310 66 3 data datum NNS web-archive-org-7310 66 4 collected collect VBN web-archive-org-7310 66 5 between between IN web-archive-org-7310 66 6 March March NNP web-archive-org-7310 66 7 and and CC web-archive-org-7310 66 8 November November NNP web-archive-org-7310 66 9 1999 1999 CD web-archive-org-7310 66 10 . . . web-archive-org-7310 67 1 For for IN web-archive-org-7310 67 2 pages page NNS web-archive-org-7310 67 3 that that WDT web-archive-org-7310 67 4 were be VBD web-archive-org-7310 67 5 downloaded download VBN web-archive-org-7310 67 6 six six CD web-archive-org-7310 67 7 times time NNS web-archive-org-7310 67 8 or or CC web-archive-org-7310 67 9 more more JJR web-archive-org-7310 67 10 , , , web-archive-org-7310 67 11 56 56 CD web-archive-org-7310 67 12 % % NN web-archive-org-7310 67 13 did do VBD web-archive-org-7310 67 14 not not RB web-archive-org-7310 67 15 change change VB web-archive-org-7310 67 16 at at RB web-archive-org-7310 67 17 all all RB web-archive-org-7310 67 18 over over IN web-archive-org-7310 67 19 the the DT web-archive-org-7310 67 20 duration duration NN web-archive-org-7310 67 21 of of IN web-archive-org-7310 67 22 the the DT web-archive-org-7310 67 23 study study NN web-archive-org-7310 67 24 ( ( -LRB- web-archive-org-7310 67 25 according accord VBG web-archive-org-7310 67 26 to to IN web-archive-org-7310 67 27 the the DT web-archive-org-7310 67 28 features feature NNS web-archive-org-7310 67 29 they -PRON- PRP web-archive-org-7310 67 30 retained retain VBD web-archive-org-7310 67 31 ) ) -RRB- web-archive-org-7310 67 32 , , , web-archive-org-7310 67 33 while while IN web-archive-org-7310 67 34 4 4 CD web-archive-org-7310 67 35 % % NN web-archive-org-7310 67 36 changed change VBD web-archive-org-7310 67 37 every every DT web-archive-org-7310 67 38 single single JJ web-archive-org-7310 67 39 time time NN web-archive-org-7310 67 40 . . . web-archive-org-7310 68 1 Our -PRON- PRP$ web-archive-org-7310 68 2 study study NN web-archive-org-7310 68 3 differs differ VBZ web-archive-org-7310 68 4 from from IN web-archive-org-7310 68 5 previous previous JJ web-archive-org-7310 68 6 studies study NNS web-archive-org-7310 68 7 in in IN web-archive-org-7310 68 8 several several JJ web-archive-org-7310 68 9 respects respect NNS web-archive-org-7310 68 10 . . . web-archive-org-7310 69 1 First first RB web-archive-org-7310 69 2 , , , web-archive-org-7310 69 3 it -PRON- PRP web-archive-org-7310 69 4 covers cover VBZ web-archive-org-7310 69 5 a a DT web-archive-org-7310 69 6 roughly roughly RB web-archive-org-7310 69 7 200 200 CD web-archive-org-7310 69 8 times time NNS web-archive-org-7310 69 9 larger large JJR web-archive-org-7310 69 10 portion portion NN web-archive-org-7310 69 11 of of IN web-archive-org-7310 69 12 the the DT web-archive-org-7310 69 13 web web NN web-archive-org-7310 69 14 ( ( -LRB- web-archive-org-7310 69 15 although although IN web-archive-org-7310 69 16 the the DT web-archive-org-7310 69 17 interval interval NN web-archive-org-7310 69 18 between between IN web-archive-org-7310 69 19 revisits revisit NNS web-archive-org-7310 69 20 is be VBZ web-archive-org-7310 69 21 seven seven CD web-archive-org-7310 69 22 times time NNS web-archive-org-7310 69 23 larger large JJR web-archive-org-7310 69 24 than than IN web-archive-org-7310 69 25 , , , web-archive-org-7310 69 26 say say UH web-archive-org-7310 69 27 , , , web-archive-org-7310 69 28 the the DT web-archive-org-7310 69 29 one one NN web-archive-org-7310 69 30 used use VBN web-archive-org-7310 69 31 by by IN web-archive-org-7310 69 32 Cho Cho NNP web-archive-org-7310 69 33 and and CC web-archive-org-7310 69 34 Garcia Garcia NNP web-archive-org-7310 69 35 - - HYPH web-archive-org-7310 69 36 Molina Molina NNP web-archive-org-7310 69 37 ) ) -RRB- web-archive-org-7310 69 38 . . . web-archive-org-7310 70 1 Second second JJ web-archive-org-7310 70 2 , , , web-archive-org-7310 70 3 we -PRON- PRP web-archive-org-7310 70 4 used use VBD web-archive-org-7310 70 5 a a DT web-archive-org-7310 70 6 different different JJ web-archive-org-7310 70 7 and and CC web-archive-org-7310 70 8 more more RBR web-archive-org-7310 70 9 fine fine RB web-archive-org-7310 70 10 - - HYPH web-archive-org-7310 70 11 grained grain VBN web-archive-org-7310 70 12 similarity similarity NN web-archive-org-7310 70 13 metric metric NN web-archive-org-7310 70 14 than than IN web-archive-org-7310 70 15 any any DT web-archive-org-7310 70 16 of of IN web-archive-org-7310 70 17 the the DT web-archive-org-7310 70 18 other other JJ web-archive-org-7310 70 19 studies study NNS web-archive-org-7310 70 20 , , , web-archive-org-7310 70 21 based base VBN web-archive-org-7310 70 22 on on IN web-archive-org-7310 70 23 syntactic syntactic JJ web-archive-org-7310 70 24 document document NN web-archive-org-7310 70 25 sketches sketch VBZ web-archive-org-7310 70 26   _SP web-archive-org-7310 70 27 [ [ -LRB- web-archive-org-7310 70 28 4 4 CD web-archive-org-7310 70 29 ] ] -RRB- web-archive-org-7310 70 30 . . . web-archive-org-7310 71 1 Third third JJ web-archive-org-7310 71 2 , , , web-archive-org-7310 71 3 we -PRON- PRP web-archive-org-7310 71 4 selected select VBD web-archive-org-7310 71 5 our -PRON- PRP$ web-archive-org-7310 71 6 pages page NNS web-archive-org-7310 71 7 based base VBN web-archive-org-7310 71 8 on on IN web-archive-org-7310 71 9 a a DT web-archive-org-7310 71 10 breadth breadth NN web-archive-org-7310 71 11 - - HYPH web-archive-org-7310 71 12 first first JJ web-archive-org-7310 71 13 crawl crawl NN web-archive-org-7310 71 14 , , , web-archive-org-7310 71 15 which which WDT web-archive-org-7310 71 16 removed remove VBD web-archive-org-7310 71 17 some some DT web-archive-org-7310 71 18 of of IN web-archive-org-7310 71 19 the the DT web-archive-org-7310 71 20 bias bias NN web-archive-org-7310 71 21 inherent inherent JJ web-archive-org-7310 71 22 in in IN web-archive-org-7310 71 23 the the DT web-archive-org-7310 71 24 other other JJ web-archive-org-7310 71 25 studies study NNS web-archive-org-7310 71 26 ( ( -LRB- web-archive-org-7310 71 27 although although IN web-archive-org-7310 71 28 breadth breadth NN web-archive-org-7310 71 29 - - HYPH web-archive-org-7310 71 30 first first JJ web-archive-org-7310 71 31 crawling crawling NN web-archive-org-7310 71 32 is be VBZ web-archive-org-7310 71 33 known know VBN web-archive-org-7310 71 34 to to TO web-archive-org-7310 71 35 be be VB web-archive-org-7310 71 36 biased bias VBN web-archive-org-7310 71 37 towards towards IN web-archive-org-7310 71 38 pages page NNS web-archive-org-7310 71 39 with with IN web-archive-org-7310 71 40 high high JJ web-archive-org-7310 71 41 PageRank PageRank NNP web-archive-org-7310 71 42   _SP web-archive-org-7310 71 43 [ [ -LRB- web-archive-org-7310 71 44 9 9 CD web-archive-org-7310 71 45 ] ] -RRB- web-archive-org-7310 71 46 ) ) -RRB- web-archive-org-7310 71 47 . . . web-archive-org-7310 72 1 Fourth fourth RB web-archive-org-7310 72 2 and and CC web-archive-org-7310 72 3 finally finally RB web-archive-org-7310 72 4 , , , web-archive-org-7310 72 5 we -PRON- PRP web-archive-org-7310 72 6 retained retain VBD web-archive-org-7310 72 7 the the DT web-archive-org-7310 72 8 full full JJ web-archive-org-7310 72 9 text text NN web-archive-org-7310 72 10 of of IN web-archive-org-7310 72 11 0.1 0.1 CD web-archive-org-7310 72 12 % % NN web-archive-org-7310 72 13 of of IN web-archive-org-7310 72 14 all all DT web-archive-org-7310 72 15 downloaded download VBN web-archive-org-7310 72 16 pages page NNS web-archive-org-7310 72 17 , , , web-archive-org-7310 72 18 a a DT web-archive-org-7310 72 19 sample sample NN web-archive-org-7310 72 20 set set NN web-archive-org-7310 72 21 that that WDT web-archive-org-7310 72 22 is be VBZ web-archive-org-7310 72 23 comparable comparable JJ web-archive-org-7310 72 24 in in IN web-archive-org-7310 72 25 size size NN web-archive-org-7310 72 26 to to IN web-archive-org-7310 72 27 the the DT web-archive-org-7310 72 28 set set NN web-archive-org-7310 72 29 of of IN web-archive-org-7310 72 30 pages page NNS web-archive-org-7310 72 31 summarized summarize VBN web-archive-org-7310 72 32 by by IN web-archive-org-7310 72 33 other other JJ web-archive-org-7310 72 34 studies study NNS web-archive-org-7310 72 35 . . . web-archive-org-7310 73 1 3 3 LS web-archive-org-7310 73 2 . . . web-archive-org-7310 74 1 EXPERIMENTAL EXPERIMENTAL NNS web-archive-org-7310 74 2 SETUP setup VBP web-archive-org-7310 74 3 Our -PRON- PRP$ web-archive-org-7310 74 4 experiment experiment NN web-archive-org-7310 74 5 can can MD web-archive-org-7310 74 6 be be VB web-archive-org-7310 74 7 divided divide VBN web-archive-org-7310 74 8 into into IN web-archive-org-7310 74 9 three three CD web-archive-org-7310 74 10 phases phase NNS web-archive-org-7310 74 11 : : : web-archive-org-7310 74 12 Collecting collect VBG web-archive-org-7310 74 13 the the DT web-archive-org-7310 74 14 data datum NNS web-archive-org-7310 74 15 through through IN web-archive-org-7310 74 16 repeated repeat VBN web-archive-org-7310 74 17 web web NN web-archive-org-7310 74 18 crawls crawl NNS web-archive-org-7310 74 19 , , , web-archive-org-7310 74 20 distilling distil VBG web-archive-org-7310 74 21 the the DT web-archive-org-7310 74 22 data datum NNS web-archive-org-7310 74 23 to to TO web-archive-org-7310 74 24 make make VB web-archive-org-7310 74 25 it -PRON- PRP web-archive-org-7310 74 26 amenable amenable JJ web-archive-org-7310 74 27 to to IN web-archive-org-7310 74 28 analysis analysis NN web-archive-org-7310 74 29 , , , web-archive-org-7310 74 30 and and CC web-archive-org-7310 74 31 mining mine VBG web-archive-org-7310 74 32 the the DT web-archive-org-7310 74 33 distillate distillate NN web-archive-org-7310 74 34 . . . web-archive-org-7310 75 1 This this DT web-archive-org-7310 75 2 section section NN web-archive-org-7310 75 3 describes describe VBZ web-archive-org-7310 75 4 each each DT web-archive-org-7310 75 5 of of IN web-archive-org-7310 75 6 the the DT web-archive-org-7310 75 7 phases phase NNS web-archive-org-7310 75 8 in in IN web-archive-org-7310 75 9 more more JJR web-archive-org-7310 75 10 detail detail NN web-archive-org-7310 75 11 . . . web-archive-org-7310 76 1 3.1 3.1 CD web-archive-org-7310 76 2       _SP web-archive-org-7310 76 3 Collecting collect VBG web-archive-org-7310 76 4 the the DT web-archive-org-7310 76 5 Data datum NNS web-archive-org-7310 76 6 Between between IN web-archive-org-7310 76 7 26 26 CD web-archive-org-7310 76 8 Nov. November NNP web-archive-org-7310 76 9 2002 2002 CD web-archive-org-7310 76 10 and and CC web-archive-org-7310 76 11 5 5 CD web-archive-org-7310 76 12 Dec. December NNP web-archive-org-7310 76 13 2002 2002 CD web-archive-org-7310 76 14 , , , web-archive-org-7310 76 15 we -PRON- PRP web-archive-org-7310 76 16 performed perform VBD web-archive-org-7310 76 17 a a DT web-archive-org-7310 76 18 large large JJ web-archive-org-7310 76 19 crawl crawl NN web-archive-org-7310 76 20 ( ( -LRB- web-archive-org-7310 76 21 " " `` web-archive-org-7310 76 22 crawl crawl NNP web-archive-org-7310 76 23 1 1 CD web-archive-org-7310 76 24 " " '' web-archive-org-7310 76 25 ) ) -RRB- web-archive-org-7310 76 26 that that WDT web-archive-org-7310 76 27 downloaded download VBD web-archive-org-7310 76 28 151 151 CD web-archive-org-7310 76 29 million million CD web-archive-org-7310 76 30 HTML html NN web-archive-org-7310 76 31 pages page NNS web-archive-org-7310 76 32 as as RB web-archive-org-7310 76 33 well well RB web-archive-org-7310 76 34 as as IN web-archive-org-7310 76 35 62 62 CD web-archive-org-7310 76 36 million million CD web-archive-org-7310 76 37 non non JJ web-archive-org-7310 76 38 - - JJ web-archive-org-7310 76 39 HTML html JJ web-archive-org-7310 76 40 pages page NNS web-archive-org-7310 76 41 , , , web-archive-org-7310 76 42 which which WDT web-archive-org-7310 76 43 we -PRON- PRP web-archive-org-7310 76 44 subsequently subsequently RB web-archive-org-7310 76 45 ignored ignore VBD web-archive-org-7310 76 46 . . . web-archive-org-7310 77 1 We -PRON- PRP web-archive-org-7310 77 2 then then RB web-archive-org-7310 77 3 attempted attempt VBD web-archive-org-7310 77 4 to to TO web-archive-org-7310 77 5 fetch fetch VB web-archive-org-7310 77 6 each each DT web-archive-org-7310 77 7 of of IN web-archive-org-7310 77 8 these these DT web-archive-org-7310 77 9 151 151 CD web-archive-org-7310 77 10 million million CD web-archive-org-7310 77 11 HTML html NN web-archive-org-7310 77 12 pages page NNS web-archive-org-7310 77 13 ten ten CD web-archive-org-7310 77 14 more more JJR web-archive-org-7310 77 15 times time NNS web-archive-org-7310 77 16 over over IN web-archive-org-7310 77 17 a a DT web-archive-org-7310 77 18 span span NN web-archive-org-7310 77 19 of of IN web-archive-org-7310 77 20 ten ten CD web-archive-org-7310 77 21 weeks week NNS web-archive-org-7310 77 22 . . . web-archive-org-7310 78 1 Naturally naturally RB web-archive-org-7310 78 2 , , , web-archive-org-7310 78 3 some some DT web-archive-org-7310 78 4 of of IN web-archive-org-7310 78 5 these these DT web-archive-org-7310 78 6 pages page NNS web-archive-org-7310 78 7 became become VBD web-archive-org-7310 78 8 either either CC web-archive-org-7310 78 9 temporarily temporarily RB web-archive-org-7310 78 10 or or CC web-archive-org-7310 78 11 permanently permanently RB web-archive-org-7310 78 12 unavailable unavailable JJ web-archive-org-7310 78 13 . . . web-archive-org-7310 79 1 Moreover moreover RB web-archive-org-7310 79 2 , , , web-archive-org-7310 79 3 we -PRON- PRP web-archive-org-7310 79 4 experienced experience VBD web-archive-org-7310 79 5 a a DT web-archive-org-7310 79 6 catastrophic catastrophic JJ web-archive-org-7310 79 7 disk disk NN web-archive-org-7310 79 8 failure failure NN web-archive-org-7310 79 9 during during IN web-archive-org-7310 79 10 the the DT web-archive-org-7310 79 11 third third JJ web-archive-org-7310 79 12 crawl crawl NN web-archive-org-7310 79 13 , , , web-archive-org-7310 79 14 causing cause VBG web-archive-org-7310 79 15 us -PRON- PRP web-archive-org-7310 79 16 to to TO web-archive-org-7310 79 17 lose lose VB web-archive-org-7310 79 18 a a DT web-archive-org-7310 79 19 quarter quarter NN web-archive-org-7310 79 20 of of IN web-archive-org-7310 79 21 the the DT web-archive-org-7310 79 22 logs log NNS web-archive-org-7310 79 23 of of IN web-archive-org-7310 79 24 that that DT web-archive-org-7310 79 25 crawl crawl NN web-archive-org-7310 79 26 . . . web-archive-org-7310 80 1 Figure figure NN web-archive-org-7310 80 2   _SP web-archive-org-7310 80 3 1 1 CD web-archive-org-7310 80 4 shows show VBZ web-archive-org-7310 80 5 the the DT web-archive-org-7310 80 6 distribution distribution NN web-archive-org-7310 80 7 of of IN web-archive-org-7310 80 8 successful successful JJ web-archive-org-7310 80 9 downloads download NNS web-archive-org-7310 80 10 . . . web-archive-org-7310 81 1 As as IN web-archive-org-7310 81 2 can can MD web-archive-org-7310 81 3 be be VB web-archive-org-7310 81 4 seen see VBN web-archive-org-7310 81 5 , , , web-archive-org-7310 81 6 we -PRON- PRP web-archive-org-7310 81 7 succeeded succeed VBD web-archive-org-7310 81 8 in in IN web-archive-org-7310 81 9 downloading download VBG web-archive-org-7310 81 10 49.2 49.2 CD web-archive-org-7310 81 11 % % NN web-archive-org-7310 81 12 of of IN web-archive-org-7310 81 13 the the DT web-archive-org-7310 81 14 pages page NNS web-archive-org-7310 81 15 all all DT web-archive-org-7310 81 16 eleven eleven CD web-archive-org-7310 81 17 times time NNS web-archive-org-7310 81 18 , , , web-archive-org-7310 81 19 and and CC web-archive-org-7310 81 20 33.6 33.6 CD web-archive-org-7310 81 21 % % NN web-archive-org-7310 81 22 ten ten CD web-archive-org-7310 81 23 times time NNS web-archive-org-7310 81 24 , , , web-archive-org-7310 81 25 leaving leave VBG web-archive-org-7310 81 26 17.2 17.2 CD web-archive-org-7310 81 27 % % NN web-archive-org-7310 81 28 of of IN web-archive-org-7310 81 29 the the DT web-archive-org-7310 81 30 pages page NNS web-archive-org-7310 81 31 that that WDT web-archive-org-7310 81 32 could could MD web-archive-org-7310 81 33 only only RB web-archive-org-7310 81 34 be be VB web-archive-org-7310 81 35 downloaded download VBN web-archive-org-7310 81 36 nine nine CD web-archive-org-7310 81 37 times time NNS web-archive-org-7310 81 38 or or CC web-archive-org-7310 81 39 fewer few JJR web-archive-org-7310 81 40 . . . web-archive-org-7310 82 1 Figure figure NN web-archive-org-7310 82 2 1 1 CD web-archive-org-7310 82 3 : : : web-archive-org-7310 82 4 Successfully successfully RB web-archive-org-7310 82 5 downloaded downloaded JJ web-archive-org-7310 82 6 versions version NNS web-archive-org-7310 82 7 of of IN web-archive-org-7310 82 8 URLs url NNS web-archive-org-7310 82 9 Our -PRON- PRP$ web-archive-org-7310 82 10 hardware hardware NN web-archive-org-7310 82 11 infrastructure infrastructure NN web-archive-org-7310 82 12 consisted consist VBN web-archive-org-7310 82 13 of of IN web-archive-org-7310 82 14 a a DT web-archive-org-7310 82 15 cluster cluster NN web-archive-org-7310 82 16 of of IN web-archive-org-7310 82 17 four four CD web-archive-org-7310 82 18 Compaq Compaq NNP web-archive-org-7310 82 19 DS20 DS20 NNP web-archive-org-7310 82 20 servers server NNS web-archive-org-7310 82 21 , , , web-archive-org-7310 82 22 each each DT web-archive-org-7310 82 23 one one NN web-archive-org-7310 82 24 equipped equip VBD web-archive-org-7310 82 25 with with IN web-archive-org-7310 82 26 a a DT web-archive-org-7310 82 27 667 667 CD web-archive-org-7310 82 28 MHz MHz NNS web-archive-org-7310 82 29 Alpha alpha NN web-archive-org-7310 82 30 processor processor NN web-archive-org-7310 82 31 , , , web-archive-org-7310 82 32 4 4 CD web-archive-org-7310 82 33 GB GB NNP web-archive-org-7310 82 34 of of IN web-archive-org-7310 82 35 RAM RAM NNP web-archive-org-7310 82 36 , , , web-archive-org-7310 82 37 648 648 CD web-archive-org-7310 82 38 GB GB NNP web-archive-org-7310 82 39 of of IN web-archive-org-7310 82 40 disk disk NN web-archive-org-7310 82 41 , , , web-archive-org-7310 82 42 and and CC web-archive-org-7310 82 43 a a DT web-archive-org-7310 82 44 fast fast JJ web-archive-org-7310 82 45 Ethernet Ethernet NNP web-archive-org-7310 82 46 network network NN web-archive-org-7310 82 47 connection connection NN web-archive-org-7310 82 48 . . . web-archive-org-7310 83 1 The the DT web-archive-org-7310 83 2 machines machine NNS web-archive-org-7310 83 3 were be VBD web-archive-org-7310 83 4 located locate VBN web-archive-org-7310 83 5 at at IN web-archive-org-7310 83 6 the the DT web-archive-org-7310 83 7 Palo Palo NNP web-archive-org-7310 83 8 Alto Alto NNP web-archive-org-7310 83 9 Internet Internet NNP web-archive-org-7310 83 10 Exchange Exchange NNP web-archive-org-7310 83 11 , , , web-archive-org-7310 83 12 a a DT web-archive-org-7310 83 13 peering peering NN web-archive-org-7310 83 14 point point NN web-archive-org-7310 83 15 for for IN web-archive-org-7310 83 16 12 12 CD web-archive-org-7310 83 17 Telcos Telcos NNPS web-archive-org-7310 83 18 and and CC web-archive-org-7310 83 19 20 20 CD web-archive-org-7310 83 20 major major JJ web-archive-org-7310 83 21 and and CC web-archive-org-7310 83 22 about about RB web-archive-org-7310 83 23 130 130 CD web-archive-org-7310 83 24 minor minor JJ web-archive-org-7310 83 25 ISPs isp NNS web-archive-org-7310 83 26 . . . web-archive-org-7310 84 1 We -PRON- PRP web-archive-org-7310 84 2 conducted conduct VBD web-archive-org-7310 84 3 these these DT web-archive-org-7310 84 4 crawls crawl NNS web-archive-org-7310 84 5 using use VBG web-archive-org-7310 84 6 the the DT web-archive-org-7310 84 7 Mercator Mercator NNP web-archive-org-7310 84 8 web web NN web-archive-org-7310 84 9 crawler crawler NN web-archive-org-7310 84 10   _SP web-archive-org-7310 84 11 [ [ -LRB- web-archive-org-7310 84 12 8 8 CD web-archive-org-7310 84 13 ] ] -RRB- web-archive-org-7310 84 14 . . . web-archive-org-7310 85 1 Mercator mercator NN web-archive-org-7310 85 2 is be VBZ web-archive-org-7310 85 3 both both CC web-archive-org-7310 85 4 fast fast JJ web-archive-org-7310 85 5 and and CC web-archive-org-7310 85 6 highly highly RB web-archive-org-7310 85 7 configurable configurable JJ web-archive-org-7310 85 8 , , , web-archive-org-7310 85 9 making make VBG web-archive-org-7310 85 10 it -PRON- PRP web-archive-org-7310 85 11 a a DT web-archive-org-7310 85 12 suitable suitable JJ web-archive-org-7310 85 13 tool tool NN web-archive-org-7310 85 14 for for IN web-archive-org-7310 85 15 our -PRON- PRP$ web-archive-org-7310 85 16 purposes purpose NNS web-archive-org-7310 85 17 . . . web-archive-org-7310 86 1 We -PRON- PRP web-archive-org-7310 86 2 seeded seed VBD web-archive-org-7310 86 3 crawl crawl NN web-archive-org-7310 86 4 1 1 CD web-archive-org-7310 86 5 with with IN web-archive-org-7310 86 6 the the DT web-archive-org-7310 86 7 Yahoo Yahoo NNP web-archive-org-7310 86 8 ! ! . web-archive-org-7310 87 1 home home NN web-archive-org-7310 87 2 page page NN web-archive-org-7310 87 3 . . . web-archive-org-7310 88 1 We -PRON- PRP web-archive-org-7310 88 2 restricted restrict VBD web-archive-org-7310 88 3 ourselves -PRON- PRP web-archive-org-7310 88 4 to to TO web-archive-org-7310 88 5 content content NN web-archive-org-7310 88 6 retrievable retrievable NN web-archive-org-7310 88 7 using use VBG web-archive-org-7310 88 8 HTTP HTTP NNP web-archive-org-7310 88 9 , , , web-archive-org-7310 88 10 ignoring ignore VBG web-archive-org-7310 88 11 HTTPS HTTPS NNP web-archive-org-7310 88 12 , , , web-archive-org-7310 88 13 FTP FTP NNP web-archive-org-7310 88 14 , , , web-archive-org-7310 88 15 Gopher Gopher NNP web-archive-org-7310 88 16 , , , web-archive-org-7310 88 17 and and CC web-archive-org-7310 88 18 the the DT web-archive-org-7310 88 19 like like JJ web-archive-org-7310 88 20 . . . web-archive-org-7310 89 1 Mercator mercator NN web-archive-org-7310 89 2 crawled crawl VBD web-archive-org-7310 89 3 using use VBG web-archive-org-7310 89 4 its -PRON- PRP$ web-archive-org-7310 89 5 standard standard JJ web-archive-org-7310 89 6 breadth breadth NN web-archive-org-7310 89 7 - - HYPH web-archive-org-7310 89 8 first first JJ web-archive-org-7310 89 9 search search NN web-archive-org-7310 89 10 web web NN web-archive-org-7310 89 11 traversal traversal NN web-archive-org-7310 89 12 strategy strategy NN web-archive-org-7310 89 13 , , , web-archive-org-7310 89 14 which which WDT web-archive-org-7310 89 15 is be VBZ web-archive-org-7310 89 16 biased bias VBN web-archive-org-7310 89 17 towards towards IN web-archive-org-7310 89 18 pages page NNS web-archive-org-7310 89 19 with with IN web-archive-org-7310 89 20 high high JJ web-archive-org-7310 89 21 PageRank PageRank NNP web-archive-org-7310 89 22   _SP web-archive-org-7310 89 23 [ [ -LRB- web-archive-org-7310 89 24 9 9 CD web-archive-org-7310 89 25 ] ] -RRB- web-archive-org-7310 89 26 . . . web-archive-org-7310 90 1 This this DT web-archive-org-7310 90 2 crawl crawl NN web-archive-org-7310 90 3 ran run VBD web-archive-org-7310 90 4 for for IN web-archive-org-7310 90 5 9 9 CD web-archive-org-7310 90 6 days day NNS web-archive-org-7310 90 7 , , , web-archive-org-7310 90 8 and and CC web-archive-org-7310 90 9 downloaded download VBD web-archive-org-7310 90 10 a a DT web-archive-org-7310 90 11 total total NN web-archive-org-7310 90 12 of of IN web-archive-org-7310 90 13 6.4 6.4 CD web-archive-org-7310 90 14 TB tb CD web-archive-org-7310 90 15 of of IN web-archive-org-7310 90 16 data datum NNS web-archive-org-7310 90 17 . . . web-archive-org-7310 91 1 As as IN web-archive-org-7310 91 2 said say VBD web-archive-org-7310 91 3 earlier early RBR web-archive-org-7310 91 4 , , , web-archive-org-7310 91 5 it -PRON- PRP web-archive-org-7310 91 6 logged log VBD web-archive-org-7310 91 7 the the DT web-archive-org-7310 91 8 URLs url NNS web-archive-org-7310 91 9 of of IN web-archive-org-7310 91 10 151 151 CD web-archive-org-7310 91 11 million million CD web-archive-org-7310 91 12 HTML html NN web-archive-org-7310 91 13 pages page NNS web-archive-org-7310 91 14 that that WDT web-archive-org-7310 91 15 were be VBD web-archive-org-7310 91 16 successfully successfully RB web-archive-org-7310 91 17 retrieved retrieve VBN web-archive-org-7310 91 18 , , , web-archive-org-7310 91 19 i.e. i.e. FW web-archive-org-7310 92 1 that that DT web-archive-org-7310 92 2 were be VBD web-archive-org-7310 92 3 returned return VBN web-archive-org-7310 92 4 with with IN web-archive-org-7310 92 5 an an DT web-archive-org-7310 92 6 HTTP http JJ web-archive-org-7310 92 7 status status NN web-archive-org-7310 92 8 code code NN web-archive-org-7310 92 9 of of IN web-archive-org-7310 92 10 200 200 CD web-archive-org-7310 92 11 and and CC web-archive-org-7310 92 12 a a DT web-archive-org-7310 92 13 content content JJ web-archive-org-7310 92 14 type type NN web-archive-org-7310 92 15 of of IN web-archive-org-7310 92 16 text text NN web-archive-org-7310 92 17 / / SYM web-archive-org-7310 92 18 html html NN web-archive-org-7310 92 19 . . . web-archive-org-7310 93 1 We -PRON- PRP web-archive-org-7310 93 2 used use VBD web-archive-org-7310 93 3 these these DT web-archive-org-7310 93 4 151 151 CD web-archive-org-7310 93 5 million million CD web-archive-org-7310 93 6 URLs url NNS web-archive-org-7310 93 7 to to TO web-archive-org-7310 93 8 seed seed VB web-archive-org-7310 93 9 the the DT web-archive-org-7310 93 10 following follow VBG web-archive-org-7310 93 11 ten ten CD web-archive-org-7310 93 12 crawls crawl NNS web-archive-org-7310 93 13 . . . web-archive-org-7310 94 1 These these DT web-archive-org-7310 94 2 crawls crawl NNS web-archive-org-7310 94 3 ran run VBD web-archive-org-7310 94 4 consecutively consecutively RB web-archive-org-7310 94 5 , , , web-archive-org-7310 94 6 starting start VBG web-archive-org-7310 94 7 on on IN web-archive-org-7310 94 8 5 5 CD web-archive-org-7310 94 9 Dec. December NNP web-archive-org-7310 94 10 2002 2002 CD web-archive-org-7310 94 11 and and CC web-archive-org-7310 94 12 ending end VBG web-archive-org-7310 94 13 on on IN web-archive-org-7310 94 14 12 12 CD web-archive-org-7310 94 15 Feb. February NNP web-archive-org-7310 94 16 2003 2003 CD web-archive-org-7310 94 17 . . . web-archive-org-7310 95 1 We -PRON- PRP web-archive-org-7310 95 2 disabled disabled JJ web-archive-org-7310 95 3 link link NN web-archive-org-7310 95 4 extraction extraction NN web-archive-org-7310 95 5 and and CC web-archive-org-7310 95 6 HTTP HTTP NNP web-archive-org-7310 95 7 redirections redirection NNS web-archive-org-7310 95 8 ; ; : web-archive-org-7310 95 9 in in IN web-archive-org-7310 95 10 other other JJ web-archive-org-7310 95 11 words word NNS web-archive-org-7310 95 12 , , , web-archive-org-7310 95 13 we -PRON- PRP web-archive-org-7310 95 14 configured configure VBD web-archive-org-7310 95 15 Mercator Mercator NNP web-archive-org-7310 95 16 to to TO web-archive-org-7310 95 17 fetch fetch VB web-archive-org-7310 95 18 only only RB web-archive-org-7310 95 19 the the DT web-archive-org-7310 95 20 pages page NNS web-archive-org-7310 95 21 in in IN web-archive-org-7310 95 22 its -PRON- PRP$ web-archive-org-7310 95 23 seed seed NN web-archive-org-7310 95 24 set set NN web-archive-org-7310 95 25 . . . web-archive-org-7310 96 1 Every every DT web-archive-org-7310 96 2 crawl crawl NN web-archive-org-7310 96 3 slowed slow VBD web-archive-org-7310 96 4 down down RP web-archive-org-7310 96 5 by by IN web-archive-org-7310 96 6 two two CD web-archive-org-7310 96 7 orders order NNS web-archive-org-7310 96 8 of of IN web-archive-org-7310 96 9 magnitude magnitude NN web-archive-org-7310 96 10 once once IN web-archive-org-7310 96 11 it -PRON- PRP web-archive-org-7310 96 12 had have VBD web-archive-org-7310 96 13 processed process VBN web-archive-org-7310 96 14 all all DT web-archive-org-7310 96 15 but but IN web-archive-org-7310 96 16 a a DT web-archive-org-7310 96 17 million million CD web-archive-org-7310 96 18 or or CC web-archive-org-7310 96 19 so so RB web-archive-org-7310 96 20 seeds seed NNS web-archive-org-7310 96 21 , , , web-archive-org-7310 96 22 because because IN web-archive-org-7310 96 23 the the DT web-archive-org-7310 96 24 remaining remaining JJ web-archive-org-7310 96 25 URLs url NNS web-archive-org-7310 96 26 all all DT web-archive-org-7310 96 27 referred refer VBN web-archive-org-7310 96 28 to to IN web-archive-org-7310 96 29 web web NN web-archive-org-7310 96 30 servers server NNS web-archive-org-7310 96 31 that that WDT web-archive-org-7310 96 32 were be VBD web-archive-org-7310 96 33 extremely extremely RB web-archive-org-7310 96 34 slow slow JJ web-archive-org-7310 96 35 to to TO web-archive-org-7310 96 36 respond respond VB web-archive-org-7310 96 37 to to IN web-archive-org-7310 96 38 us -PRON- PRP web-archive-org-7310 96 39 during during IN web-archive-org-7310 96 40 that that DT web-archive-org-7310 96 41 crawl crawl NN web-archive-org-7310 96 42 , , , web-archive-org-7310 96 43 and and CC web-archive-org-7310 96 44 because because IN web-archive-org-7310 96 45 Mercator Mercator NNP web-archive-org-7310 96 46 's 's POS web-archive-org-7310 96 47 politeness politeness NN web-archive-org-7310 96 48 policies policy NNS web-archive-org-7310 96 49 cause cause VBP web-archive-org-7310 96 50 it -PRON- PRP web-archive-org-7310 96 51 to to TO web-archive-org-7310 96 52 space space VB web-archive-org-7310 96 53 out out RP web-archive-org-7310 96 54 requests request NNS web-archive-org-7310 96 55 to to IN web-archive-org-7310 96 56 the the DT web-archive-org-7310 96 57 same same JJ web-archive-org-7310 96 58 host host NN web-archive-org-7310 96 59 proportional proportional JJ web-archive-org-7310 96 60 to to IN web-archive-org-7310 96 61 the the DT web-archive-org-7310 96 62 delay delay NN web-archive-org-7310 96 63 of of IN web-archive-org-7310 96 64 the the DT web-archive-org-7310 96 65 previous previous JJ web-archive-org-7310 96 66 response response NN web-archive-org-7310 96 67 from from IN web-archive-org-7310 96 68 that that DT web-archive-org-7310 96 69 host host NN web-archive-org-7310 96 70 . . . web-archive-org-7310 97 1 In in IN web-archive-org-7310 97 2 order order NN web-archive-org-7310 97 3 to to TO web-archive-org-7310 97 4 keep keep VB web-archive-org-7310 97 5 the the DT web-archive-org-7310 97 6 overall overall JJ web-archive-org-7310 97 7 duration duration NN web-archive-org-7310 97 8 of of IN web-archive-org-7310 97 9 each each DT web-archive-org-7310 97 10 crawl crawl NN web-archive-org-7310 97 11 within within IN web-archive-org-7310 97 12 reasonable reasonable JJ web-archive-org-7310 97 13 bounds bound NNS web-archive-org-7310 97 14 , , , web-archive-org-7310 97 15 we -PRON- PRP web-archive-org-7310 97 16 terminated terminate VBD web-archive-org-7310 97 17 the the DT web-archive-org-7310 97 18 crawls crawl NNS web-archive-org-7310 97 19 after after IN web-archive-org-7310 97 20 this this DT web-archive-org-7310 97 21 happened happen VBD web-archive-org-7310 97 22 , , , web-archive-org-7310 97 23 typically typically RB web-archive-org-7310 97 24 on on IN web-archive-org-7310 97 25 the the DT web-archive-org-7310 97 26 sixth sixth JJ web-archive-org-7310 97 27 day day NN web-archive-org-7310 97 28 of of IN web-archive-org-7310 97 29 the the DT web-archive-org-7310 97 30 crawl crawl NN web-archive-org-7310 97 31 , , , web-archive-org-7310 97 32 and and CC web-archive-org-7310 97 33 started start VBD web-archive-org-7310 97 34 the the DT web-archive-org-7310 97 35 next next JJ web-archive-org-7310 97 36 crawl crawl NN web-archive-org-7310 97 37 a a DT web-archive-org-7310 97 38 week week NN web-archive-org-7310 97 39 after after IN web-archive-org-7310 97 40 the the DT web-archive-org-7310 97 41 start start NN web-archive-org-7310 97 42 of of IN web-archive-org-7310 97 43 the the DT web-archive-org-7310 97 44 preceding precede VBG web-archive-org-7310 97 45 one one CD web-archive-org-7310 97 46 . . . web-archive-org-7310 98 1 For for IN web-archive-org-7310 98 2 all all DT web-archive-org-7310 98 3 eleven eleven CD web-archive-org-7310 98 4 crawls crawl NNS web-archive-org-7310 98 5 , , , web-archive-org-7310 98 6 we -PRON- PRP web-archive-org-7310 98 7 provided provide VBD web-archive-org-7310 98 8 Mercator mercator NN web-archive-org-7310 98 9 with with IN web-archive-org-7310 98 10 two two CD web-archive-org-7310 98 11 new new JJ web-archive-org-7310 98 12 processing processing NN web-archive-org-7310 98 13 modules module NNS web-archive-org-7310 98 14 : : : web-archive-org-7310 98 15 a a DT web-archive-org-7310 98 16 module module NN web-archive-org-7310 98 17 that that WDT web-archive-org-7310 98 18 recorded record VBD web-archive-org-7310 98 19 a a DT web-archive-org-7310 98 20 checksum checksum NN web-archive-org-7310 98 21 and and CC web-archive-org-7310 98 22 a a DT web-archive-org-7310 98 23 fixed fix VBN web-archive-org-7310 98 24 - - HYPH web-archive-org-7310 98 25 size size NN web-archive-org-7310 98 26 feature feature NN web-archive-org-7310 98 27 vector vector NN web-archive-org-7310 98 28 plus plus CC web-archive-org-7310 98 29 some some DT web-archive-org-7310 98 30 ancillary ancillary JJ web-archive-org-7310 98 31 data datum NNS web-archive-org-7310 98 32 for for IN web-archive-org-7310 98 33 each each DT web-archive-org-7310 98 34 page page NN web-archive-org-7310 98 35 , , , web-archive-org-7310 98 36 and and CC web-archive-org-7310 98 37 a a DT web-archive-org-7310 98 38 second second JJ web-archive-org-7310 98 39 module module NN web-archive-org-7310 98 40 that that WDT web-archive-org-7310 98 41 selected select VBD web-archive-org-7310 98 42 0.1 0.1 CD web-archive-org-7310 98 43 % % NN web-archive-org-7310 98 44 of of IN web-archive-org-7310 98 45 all all DT web-archive-org-7310 98 46 pages page NNS web-archive-org-7310 98 47 and and CC web-archive-org-7310 98 48 saved save VBD web-archive-org-7310 98 49 them -PRON- PRP web-archive-org-7310 98 50 to to IN web-archive-org-7310 98 51 disk disk NN web-archive-org-7310 98 52 . . . web-archive-org-7310 99 1 We -PRON- PRP web-archive-org-7310 99 2 computed compute VBD web-archive-org-7310 99 3 the the DT web-archive-org-7310 99 4 feature feature NN web-archive-org-7310 99 5 vectors vector NNS web-archive-org-7310 99 6 using use VBG web-archive-org-7310 99 7 a a DT web-archive-org-7310 99 8 modified modify VBN web-archive-org-7310 99 9 version version NN web-archive-org-7310 99 10 of of IN web-archive-org-7310 99 11 the the DT web-archive-org-7310 99 12 document document NN web-archive-org-7310 99 13 shingling shingling NN web-archive-org-7310 99 14 technique technique NN web-archive-org-7310 99 15 due due JJ web-archive-org-7310 99 16 to to IN web-archive-org-7310 99 17 Broder Broder NNP web-archive-org-7310 99 18 et et NNP web-archive-org-7310 99 19 al al NNP web-archive-org-7310 99 20 . . . web-archive-org-7310 100 1 [ [ -LRB- web-archive-org-7310 100 2 4 4 CD web-archive-org-7310 100 3 ] ] -RRB- web-archive-org-7310 100 4 , , , web-archive-org-7310 100 5 which which WDT web-archive-org-7310 100 6 uses use VBZ web-archive-org-7310 100 7 a a DT web-archive-org-7310 100 8 metric metric NN web-archive-org-7310 100 9 of of IN web-archive-org-7310 100 10 document document NN web-archive-org-7310 100 11 similarity similarity NN web-archive-org-7310 100 12 based base VBN web-archive-org-7310 100 13 on on IN web-archive-org-7310 100 14 syntactic syntactic JJ web-archive-org-7310 100 15 properties property NNS web-archive-org-7310 100 16 of of IN web-archive-org-7310 100 17 the the DT web-archive-org-7310 100 18 document document NN web-archive-org-7310 100 19 . . . web-archive-org-7310 101 1 This this DT web-archive-org-7310 101 2 similarity similarity NN web-archive-org-7310 101 3 metric metric NN web-archive-org-7310 101 4 is be VBZ web-archive-org-7310 101 5 applicable applicable JJ web-archive-org-7310 101 6 to to IN web-archive-org-7310 101 7 any any DT web-archive-org-7310 101 8 kind kind NN web-archive-org-7310 101 9 of of IN web-archive-org-7310 101 10 document document NN web-archive-org-7310 101 11 that that WDT web-archive-org-7310 101 12 consists consist VBZ web-archive-org-7310 101 13 of of IN web-archive-org-7310 101 14 an an DT web-archive-org-7310 101 15 ordered order VBN web-archive-org-7310 101 16 sequence sequence NN web-archive-org-7310 101 17 of of IN web-archive-org-7310 101 18 discrete discrete JJ web-archive-org-7310 101 19 features feature NNS web-archive-org-7310 101 20 . . . web-archive-org-7310 102 1 The the DT web-archive-org-7310 102 2 following follow VBG web-archive-org-7310 102 3 description description NN web-archive-org-7310 102 4 assumes assume VBZ web-archive-org-7310 102 5 that that IN web-archive-org-7310 102 6 the the DT web-archive-org-7310 102 7 features feature NNS web-archive-org-7310 102 8 are be VBP web-archive-org-7310 102 9 words word NNS web-archive-org-7310 102 10 , , , web-archive-org-7310 102 11 i.e. i.e. FW web-archive-org-7310 103 1 that that IN web-archive-org-7310 103 2 a a DT web-archive-org-7310 103 3 document document NN web-archive-org-7310 103 4 is be VBZ web-archive-org-7310 103 5 an an DT web-archive-org-7310 103 6 ordered order VBN web-archive-org-7310 103 7 sequence sequence NN web-archive-org-7310 103 8 of of IN web-archive-org-7310 103 9 words word NNS web-archive-org-7310 103 10 . . . web-archive-org-7310 104 1 In in IN web-archive-org-7310 104 2 order order NN web-archive-org-7310 104 3 to to TO web-archive-org-7310 104 4 compare compare VB web-archive-org-7310 104 5 two two CD web-archive-org-7310 104 6 documents document NNS web-archive-org-7310 104 7 , , , web-archive-org-7310 104 8 we -PRON- PRP web-archive-org-7310 104 9 map map VBP web-archive-org-7310 104 10 each each DT web-archive-org-7310 104 11 document document NN web-archive-org-7310 104 12 into into IN web-archive-org-7310 104 13 a a DT web-archive-org-7310 104 14 set set NN web-archive-org-7310 104 15 of of IN web-archive-org-7310 104 16 k k NN web-archive-org-7310 104 17 - - HYPH web-archive-org-7310 104 18 word word NN web-archive-org-7310 104 19 subsequences subsequence NNS web-archive-org-7310 104 20 ( ( -LRB- web-archive-org-7310 104 21 groups group NNS web-archive-org-7310 104 22 of of IN web-archive-org-7310 104 23 adjacent adjacent JJ web-archive-org-7310 104 24 words word NNS web-archive-org-7310 104 25 or or CC web-archive-org-7310 104 26 " " `` web-archive-org-7310 104 27 shingles shingle NNS web-archive-org-7310 104 28 " " '' web-archive-org-7310 104 29 ) ) -RRB- web-archive-org-7310 104 30 , , , web-archive-org-7310 104 31 wrapping wrap VBG web-archive-org-7310 104 32 at at IN web-archive-org-7310 104 33 the the DT web-archive-org-7310 104 34 end end NN web-archive-org-7310 104 35 of of IN web-archive-org-7310 104 36 the the DT web-archive-org-7310 104 37 document document NN web-archive-org-7310 104 38 , , , web-archive-org-7310 104 39 so so IN web-archive-org-7310 104 40 that that IN web-archive-org-7310 104 41 every every DT web-archive-org-7310 104 42 word word NN web-archive-org-7310 104 43 in in IN web-archive-org-7310 104 44 the the DT web-archive-org-7310 104 45 document document NN web-archive-org-7310 104 46 starts start VBZ web-archive-org-7310 104 47 a a DT web-archive-org-7310 104 48 shingle shingle NN web-archive-org-7310 104 49 . . . web-archive-org-7310 105 1 Two two CD web-archive-org-7310 105 2 documents document NNS web-archive-org-7310 105 3 are be VBP web-archive-org-7310 105 4 considered consider VBN web-archive-org-7310 105 5 to to TO web-archive-org-7310 105 6 be be VB web-archive-org-7310 105 7 identical identical JJ web-archive-org-7310 105 8 if if IN web-archive-org-7310 105 9 they -PRON- PRP web-archive-org-7310 105 10 map map VBP web-archive-org-7310 105 11 to to IN web-archive-org-7310 105 12 the the DT web-archive-org-7310 105 13 same same JJ web-archive-org-7310 105 14 set set NN web-archive-org-7310 105 15 of of IN web-archive-org-7310 105 16 shingles;1 shingles;1 NNP web-archive-org-7310 105 17 they -PRON- PRP web-archive-org-7310 105 18 are be VBP web-archive-org-7310 105 19 considered consider VBN web-archive-org-7310 105 20 to to TO web-archive-org-7310 105 21 be be VB web-archive-org-7310 105 22 similar similar JJ web-archive-org-7310 105 23 if if IN web-archive-org-7310 105 24 they -PRON- PRP web-archive-org-7310 105 25 map map VBP web-archive-org-7310 105 26 to to IN web-archive-org-7310 105 27 similar similar JJ web-archive-org-7310 105 28 sets set NNS web-archive-org-7310 105 29 of of IN web-archive-org-7310 105 30 shingles shingle NNS web-archive-org-7310 105 31 . . . web-archive-org-7310 106 1 Quantitatively quantitatively RB web-archive-org-7310 106 2 , , , web-archive-org-7310 106 3 the the DT web-archive-org-7310 106 4 similarity similarity NN web-archive-org-7310 106 5 of of IN web-archive-org-7310 106 6 two two CD web-archive-org-7310 106 7 documents document NNS web-archive-org-7310 106 8 is be VBZ web-archive-org-7310 106 9 defined define VBN web-archive-org-7310 106 10 to to TO web-archive-org-7310 106 11 be be VB web-archive-org-7310 106 12 the the DT web-archive-org-7310 106 13 number number NN web-archive-org-7310 106 14 of of IN web-archive-org-7310 106 15 distinct distinct JJ web-archive-org-7310 106 16 shingles shingle NNS web-archive-org-7310 106 17 appearing appear VBG web-archive-org-7310 106 18 in in IN web-archive-org-7310 106 19 both both DT web-archive-org-7310 106 20 documents document NNS web-archive-org-7310 106 21 divided divide VBN web-archive-org-7310 106 22 by by IN web-archive-org-7310 106 23 the the DT web-archive-org-7310 106 24 total total JJ web-archive-org-7310 106 25 number number NN web-archive-org-7310 106 26 of of IN web-archive-org-7310 106 27 distinct distinct JJ web-archive-org-7310 106 28 shingles shingle NNS web-archive-org-7310 106 29 . . . web-archive-org-7310 107 1 This this DT web-archive-org-7310 107 2 means mean VBZ web-archive-org-7310 107 3 that that IN web-archive-org-7310 107 4 two two CD web-archive-org-7310 107 5 identical identical JJ web-archive-org-7310 107 6 documents document NNS web-archive-org-7310 107 7 have have VBP web-archive-org-7310 107 8 similarity similarity NN web-archive-org-7310 107 9 1 1 CD web-archive-org-7310 107 10 , , , web-archive-org-7310 107 11 while while IN web-archive-org-7310 107 12 two two CD web-archive-org-7310 107 13 documents document NNS web-archive-org-7310 107 14 that that WDT web-archive-org-7310 107 15 have have VBP web-archive-org-7310 107 16 no no DT web-archive-org-7310 107 17 shingle shingle NN web-archive-org-7310 107 18 in in IN web-archive-org-7310 107 19 common common JJ web-archive-org-7310 107 20 have have VB web-archive-org-7310 107 21 similarity similarity NN web-archive-org-7310 107 22 0 0 CD web-archive-org-7310 107 23 . . . web-archive-org-7310 108 1 Note note VB web-archive-org-7310 108 2 that that IN web-archive-org-7310 108 3 the the DT web-archive-org-7310 108 4 value value NN web-archive-org-7310 108 5 of of IN web-archive-org-7310 108 6 k k NN web-archive-org-7310 108 7 parameterizes parameterize NNS web-archive-org-7310 108 8 how how WRB web-archive-org-7310 108 9 sensitive sensitive JJ web-archive-org-7310 108 10 this this DT web-archive-org-7310 108 11 metric metric NN web-archive-org-7310 108 12 is be VBZ web-archive-org-7310 108 13 . . . web-archive-org-7310 109 1 Changing change VBG web-archive-org-7310 109 2 the the DT web-archive-org-7310 109 3 word word NN web-archive-org-7310 109 4 " " `` web-archive-org-7310 109 5 kumquat kumquat NN web-archive-org-7310 109 6 " " '' web-archive-org-7310 109 7 to to TO web-archive-org-7310 109 8 " " `` web-archive-org-7310 109 9 persimmon persimmon VB web-archive-org-7310 109 10 " " '' web-archive-org-7310 109 11 in in IN web-archive-org-7310 109 12 an an DT web-archive-org-7310 109 13 n n JJ web-archive-org-7310 109 14 - - HYPH web-archive-org-7310 109 15 word word NN web-archive-org-7310 109 16 document document NN web-archive-org-7310 109 17 ( ( -LRB- web-archive-org-7310 109 18 assuming assume VBG web-archive-org-7310 109 19 that that IN web-archive-org-7310 109 20 " " `` web-archive-org-7310 109 21 kumquat kumquat NN web-archive-org-7310 109 22 " " '' web-archive-org-7310 109 23 and and CC web-archive-org-7310 109 24 " " `` web-archive-org-7310 109 25 persimmon persimmon NN web-archive-org-7310 109 26 " " '' web-archive-org-7310 109 27 occur occur VBP web-archive-org-7310 109 28 nowhere nowhere RB web-archive-org-7310 109 29 else else RB web-archive-org-7310 109 30 in in IN web-archive-org-7310 109 31 the the DT web-archive-org-7310 109 32 document document NN web-archive-org-7310 109 33 ) ) -RRB- web-archive-org-7310 109 34 results result VBZ web-archive-org-7310 109 35 in in IN web-archive-org-7310 109 36 a a DT web-archive-org-7310 109 37 similarity similarity NN web-archive-org-7310 109 38 of of IN web-archive-org-7310 109 39 [ [ -LRB- web-archive-org-7310 109 40 ( ( -LRB- web-archive-org-7310 109 41 n n NNP web-archive-org-7310 109 42 - - HYPH web-archive-org-7310 109 43 k)/(n+k k)/(n+k NNP web-archive-org-7310 109 44 ) ) -RRB- web-archive-org-7310 109 45 ] ] -RRB- web-archive-org-7310 109 46 between between IN web-archive-org-7310 109 47 the the DT web-archive-org-7310 109 48 original original JJ web-archive-org-7310 109 49 and and CC web-archive-org-7310 109 50 the the DT web-archive-org-7310 109 51 modified modify VBN web-archive-org-7310 109 52 document document NN web-archive-org-7310 109 53 . . . web-archive-org-7310 110 1 This this DT web-archive-org-7310 110 2 means mean VBZ web-archive-org-7310 110 3 that that IN web-archive-org-7310 110 4 one one PRP web-archive-org-7310 110 5 should should MD web-archive-org-7310 110 6 not not RB web-archive-org-7310 110 7 choose choose VB web-archive-org-7310 110 8 k k NNP web-archive-org-7310 110 9 to to TO web-archive-org-7310 110 10 be be VB web-archive-org-7310 110 11 too too RB web-archive-org-7310 110 12 large large JJ web-archive-org-7310 110 13 , , , web-archive-org-7310 110 14 lest l JJS web-archive-org-7310 110 15 even even RB web-archive-org-7310 110 16 small small JJ web-archive-org-7310 110 17 changes change NNS web-archive-org-7310 110 18 result result VBP web-archive-org-7310 110 19 in in IN web-archive-org-7310 110 20 low low JJ web-archive-org-7310 110 21 similarity similarity NN web-archive-org-7310 110 22 . . . web-archive-org-7310 111 1 On on IN web-archive-org-7310 111 2 the the DT web-archive-org-7310 111 3 other other JJ web-archive-org-7310 111 4 hand hand NN web-archive-org-7310 111 5 , , , web-archive-org-7310 111 6 neither neither CC web-archive-org-7310 111 7 should should MD web-archive-org-7310 111 8 one one PRP web-archive-org-7310 111 9 choose choose VB web-archive-org-7310 111 10 k k NNP web-archive-org-7310 111 11 to to TO web-archive-org-7310 111 12 be be VB web-archive-org-7310 111 13 too too RB web-archive-org-7310 111 14 small small JJ web-archive-org-7310 111 15 . . . web-archive-org-7310 112 1 As as IN web-archive-org-7310 112 2 an an DT web-archive-org-7310 112 3 extreme extreme JJ web-archive-org-7310 112 4 example example NN web-archive-org-7310 112 5 , , , web-archive-org-7310 112 6 setting set VBG web-archive-org-7310 112 7 k k NN web-archive-org-7310 112 8 to to IN web-archive-org-7310 112 9 1 1 CD web-archive-org-7310 112 10 results result NNS web-archive-org-7310 112 11 in in IN web-archive-org-7310 112 12 a a DT web-archive-org-7310 112 13 comparison comparison NN web-archive-org-7310 112 14 of of IN web-archive-org-7310 112 15 the the DT web-archive-org-7310 112 16 lexicon lexicon NN web-archive-org-7310 112 17 of of IN web-archive-org-7310 112 18 two two CD web-archive-org-7310 112 19 documents document NNS web-archive-org-7310 112 20 , , , web-archive-org-7310 112 21 making make VBG web-archive-org-7310 112 22 the the DT web-archive-org-7310 112 23 metric metric NN web-archive-org-7310 112 24 completely completely RB web-archive-org-7310 112 25 insensitive insensitive JJ web-archive-org-7310 112 26 to to IN web-archive-org-7310 112 27 word word NN web-archive-org-7310 112 28 ordering ordering NN web-archive-org-7310 112 29 . . . web-archive-org-7310 113 1 Our -PRON- PRP$ web-archive-org-7310 113 2 feature feature NN web-archive-org-7310 113 3 vector vector NN web-archive-org-7310 113 4 extraction extraction NN web-archive-org-7310 113 5 module module NN web-archive-org-7310 113 6 substitutes substitute VBZ web-archive-org-7310 113 7 HTML html NN web-archive-org-7310 113 8 markup markup NN web-archive-org-7310 113 9 by by IN web-archive-org-7310 113 10 whitespace whitespace NN web-archive-org-7310 113 11 , , , web-archive-org-7310 113 12 and and CC web-archive-org-7310 113 13 then then RB web-archive-org-7310 113 14 segments segment VBZ web-archive-org-7310 113 15 the the DT web-archive-org-7310 113 16 document document NN web-archive-org-7310 113 17 into into IN web-archive-org-7310 113 18 5-word 5-word CD web-archive-org-7310 113 19 shingles shingle NNS web-archive-org-7310 113 20 , , , web-archive-org-7310 113 21 where where WRB web-archive-org-7310 113 22 each each DT web-archive-org-7310 113 23 word word NN web-archive-org-7310 113 24 is be VBZ web-archive-org-7310 113 25 an an DT web-archive-org-7310 113 26 uninterrupted uninterrupted JJ web-archive-org-7310 113 27 sequence sequence NN web-archive-org-7310 113 28 of of IN web-archive-org-7310 113 29 alphanumeric alphanumeric JJ web-archive-org-7310 113 30 characters character NNS web-archive-org-7310 113 31 . . . web-archive-org-7310 114 1 Next next RB web-archive-org-7310 114 2 , , , web-archive-org-7310 114 3 it -PRON- PRP web-archive-org-7310 114 4 computes compute VBZ web-archive-org-7310 114 5 a a DT web-archive-org-7310 114 6 64-bit 64-bit CD web-archive-org-7310 114 7 checksum checksum NN web-archive-org-7310 114 8 of of IN web-archive-org-7310 114 9 each each DT web-archive-org-7310 114 10 shingle shingle NN web-archive-org-7310 114 11 , , , web-archive-org-7310 114 12 using use VBG web-archive-org-7310 114 13 Rabin Rabin NNP web-archive-org-7310 114 14 's 's POS web-archive-org-7310 114 15 fingerprinting fingerprint VBG web-archive-org-7310 114 16 algorithm algorithm NN web-archive-org-7310 114 17   _SP web-archive-org-7310 114 18 [ [ -LRB- web-archive-org-7310 114 19 3,10 3,10 CD web-archive-org-7310 114 20 ] ] -RRB- web-archive-org-7310 114 21 . . . web-archive-org-7310 115 1 We -PRON- PRP web-archive-org-7310 115 2 call call VBP web-archive-org-7310 115 3 these these DT web-archive-org-7310 115 4 fingerprints fingerprint NNS web-archive-org-7310 115 5 the the DT web-archive-org-7310 115 6 " " `` web-archive-org-7310 115 7 pre pre JJ web-archive-org-7310 115 8 - - NNS web-archive-org-7310 115 9 images image NNS web-archive-org-7310 115 10 " " '' web-archive-org-7310 115 11 . . . web-archive-org-7310 116 1 Next next RB web-archive-org-7310 116 2 , , , web-archive-org-7310 116 3 the the DT web-archive-org-7310 116 4 module module NN web-archive-org-7310 116 5 applies apply VBZ web-archive-org-7310 116 6 84 84 CD web-archive-org-7310 116 7 different different JJ web-archive-org-7310 116 8 ( ( -LRB- web-archive-org-7310 116 9 randomly randomly RB web-archive-org-7310 116 10 selected select VBN web-archive-org-7310 116 11 but but CC web-archive-org-7310 116 12 fixed fix VBN web-archive-org-7310 116 13 thereafter thereafter RB web-archive-org-7310 116 14 ) ) -RRB- web-archive-org-7310 116 15 one one CD web-archive-org-7310 116 16 - - HYPH web-archive-org-7310 116 17 to to IN web-archive-org-7310 116 18 - - HYPH web-archive-org-7310 116 19 one one CD web-archive-org-7310 116 20 functions function NNS web-archive-org-7310 116 21 to to IN web-archive-org-7310 116 22 each each DT web-archive-org-7310 116 23 pre pre NN web-archive-org-7310 116 24 - - NN web-archive-org-7310 116 25 image image NN web-archive-org-7310 116 26 . . . web-archive-org-7310 117 1 For for IN web-archive-org-7310 117 2 each each DT web-archive-org-7310 117 3 function function NN web-archive-org-7310 117 4 , , , web-archive-org-7310 117 5 we -PRON- PRP web-archive-org-7310 117 6 retain retain VBP web-archive-org-7310 117 7 the the DT web-archive-org-7310 117 8 pre pre NN web-archive-org-7310 117 9 - - NN web-archive-org-7310 117 10 image image NN web-archive-org-7310 117 11 which which WDT web-archive-org-7310 117 12 results result VBZ web-archive-org-7310 117 13 in in IN web-archive-org-7310 117 14 the the DT web-archive-org-7310 117 15 numerically numerically RB web-archive-org-7310 117 16 smallest small JJS web-archive-org-7310 117 17 image image NN web-archive-org-7310 117 18 . . . web-archive-org-7310 118 1 This this DT web-archive-org-7310 118 2 results result VBZ web-archive-org-7310 118 3 in in IN web-archive-org-7310 118 4 a a DT web-archive-org-7310 118 5 vector vector NN web-archive-org-7310 118 6 of of IN web-archive-org-7310 118 7 84 84 CD web-archive-org-7310 118 8 pre pre NNS web-archive-org-7310 118 9 - - NNS web-archive-org-7310 118 10 images image NNS web-archive-org-7310 118 11 , , , web-archive-org-7310 118 12 which which WDT web-archive-org-7310 118 13 is be VBZ web-archive-org-7310 118 14 the the DT web-archive-org-7310 118 15 desired desire VBN web-archive-org-7310 118 16 feature feature NN web-archive-org-7310 118 17 vector vector NN web-archive-org-7310 118 18 . . . web-archive-org-7310 119 1 If if IN web-archive-org-7310 119 2 the the DT web-archive-org-7310 119 3 one one CD web-archive-org-7310 119 4 - - HYPH web-archive-org-7310 119 5 to to IN web-archive-org-7310 119 6 - - HYPH web-archive-org-7310 119 7 one one CD web-archive-org-7310 119 8 functions function NNS web-archive-org-7310 119 9 are be VBP web-archive-org-7310 119 10 chosen choose VBN web-archive-org-7310 119 11 randomly,2 randomly,2 NNPS web-archive-org-7310 119 12 then then RB web-archive-org-7310 119 13 given give VBN web-archive-org-7310 119 14 the the DT web-archive-org-7310 119 15 feature feature NN web-archive-org-7310 119 16 vectors vector NNS web-archive-org-7310 119 17 of of IN web-archive-org-7310 119 18 two two CD web-archive-org-7310 119 19 documents document NNS web-archive-org-7310 119 20 , , , web-archive-org-7310 119 21 two two CD web-archive-org-7310 119 22 corresponding corresponding JJ web-archive-org-7310 119 23 elements element NNS web-archive-org-7310 119 24 of of IN web-archive-org-7310 119 25 the the DT web-archive-org-7310 119 26 vectors vector NNS web-archive-org-7310 119 27 are be VBP web-archive-org-7310 119 28 identical identical JJ web-archive-org-7310 119 29 with with IN web-archive-org-7310 119 30 probability probability NN web-archive-org-7310 119 31 equal equal JJ web-archive-org-7310 119 32 to to IN web-archive-org-7310 119 33 the the DT web-archive-org-7310 119 34 similarity similarity NN web-archive-org-7310 119 35 of of IN web-archive-org-7310 119 36 the the DT web-archive-org-7310 119 37 documents document NNS web-archive-org-7310 119 38 . . . web-archive-org-7310 120 1 The the DT web-archive-org-7310 120 2 feature feature NN web-archive-org-7310 120 3 vector vector NN web-archive-org-7310 120 4 extraction extraction NN web-archive-org-7310 120 5 module module NN web-archive-org-7310 120 6 logs log VBZ web-archive-org-7310 120 7 the the DT web-archive-org-7310 120 8 feature feature NN web-archive-org-7310 120 9 vector vector NN web-archive-org-7310 120 10 of of IN web-archive-org-7310 120 11 each each DT web-archive-org-7310 120 12 document document NN web-archive-org-7310 120 13 , , , web-archive-org-7310 120 14 together together RB web-archive-org-7310 120 15 with with IN web-archive-org-7310 120 16 a a DT web-archive-org-7310 120 17 checksum checksum NN web-archive-org-7310 120 18 of of IN web-archive-org-7310 120 19 its -PRON- PRP$ web-archive-org-7310 120 20 raw raw JJ web-archive-org-7310 120 21 content content NN web-archive-org-7310 120 22 , , , web-archive-org-7310 120 23 the the DT web-archive-org-7310 120 24 start start NN web-archive-org-7310 120 25 time time NN web-archive-org-7310 120 26 and and CC web-archive-org-7310 120 27 the the DT web-archive-org-7310 120 28 duration duration NN web-archive-org-7310 120 29 of of IN web-archive-org-7310 120 30 its -PRON- PRP$ web-archive-org-7310 120 31 retrieval retrieval NN web-archive-org-7310 120 32 , , , web-archive-org-7310 120 33 the the DT web-archive-org-7310 120 34 HTTP http JJ web-archive-org-7310 120 35 status status NN web-archive-org-7310 120 36 code code NN web-archive-org-7310 120 37 ( ( -LRB- web-archive-org-7310 120 38 or or CC web-archive-org-7310 120 39 an an DT web-archive-org-7310 120 40 error error NN web-archive-org-7310 120 41 code code NN web-archive-org-7310 120 42 indicating indicate VBG web-archive-org-7310 120 43 a a DT web-archive-org-7310 120 44 TCP TCP NNP web-archive-org-7310 120 45 error error NN web-archive-org-7310 120 46 or or CC web-archive-org-7310 120 47 a a DT web-archive-org-7310 120 48 robot robot NN web-archive-org-7310 120 49 exclusion exclusion NN web-archive-org-7310 120 50 ) ) -RRB- web-archive-org-7310 120 51 , , , web-archive-org-7310 120 52 the the DT web-archive-org-7310 120 53 document document NN web-archive-org-7310 120 54 's 's POS web-archive-org-7310 120 55 length length NN web-archive-org-7310 120 56 , , , web-archive-org-7310 120 57 the the DT web-archive-org-7310 120 58 number number NN web-archive-org-7310 120 59 of of IN web-archive-org-7310 120 60 non non JJ web-archive-org-7310 120 61 - - JJ web-archive-org-7310 120 62 markup markup JJ web-archive-org-7310 120 63 words word NNS web-archive-org-7310 120 64 , , , web-archive-org-7310 120 65 and and CC web-archive-org-7310 120 66 the the DT web-archive-org-7310 120 67 URL url NN web-archive-org-7310 120 68 . . . web-archive-org-7310 121 1 If if IN web-archive-org-7310 121 2 the the DT web-archive-org-7310 121 3 document document NN web-archive-org-7310 121 4 can can MD web-archive-org-7310 121 5 not not RB web-archive-org-7310 121 6 be be VB web-archive-org-7310 121 7 downloaded download VBN web-archive-org-7310 121 8 successfully successfully RB web-archive-org-7310 121 9 or or CC web-archive-org-7310 121 10 does do VBZ web-archive-org-7310 121 11 not not RB web-archive-org-7310 121 12 result result VB web-archive-org-7310 121 13 in in IN web-archive-org-7310 121 14 an an DT web-archive-org-7310 121 15 HTML html NN web-archive-org-7310 121 16 document,3 document,3 VB web-archive-org-7310 121 17 the the DT web-archive-org-7310 121 18 module module NN web-archive-org-7310 121 19 logs log VBZ web-archive-org-7310 121 20 the the DT web-archive-org-7310 121 21 URL url NN web-archive-org-7310 121 22 and and CC web-archive-org-7310 121 23 special special JJ web-archive-org-7310 121 24 values value NNS web-archive-org-7310 121 25 for for IN web-archive-org-7310 121 26 everything everything NN web-archive-org-7310 121 27 else else RB web-archive-org-7310 121 28 . . . web-archive-org-7310 122 1 URLs url NNS web-archive-org-7310 122 2 that that WDT web-archive-org-7310 122 3 were be VBD web-archive-org-7310 122 4 not not RB web-archive-org-7310 122 5 downloaded download VBN web-archive-org-7310 122 6 because because IN web-archive-org-7310 122 7 the the DT web-archive-org-7310 122 8 crawl crawl NN web-archive-org-7310 122 9 was be VBD web-archive-org-7310 122 10 terminated terminate VBN web-archive-org-7310 122 11 before before IN web-archive-org-7310 122 12 it -PRON- PRP web-archive-org-7310 122 13 was be VBD web-archive-org-7310 122 14 complete complete JJ web-archive-org-7310 122 15 are be VBP web-archive-org-7310 122 16 treated treat VBN web-archive-org-7310 122 17 in in IN web-archive-org-7310 122 18 a a DT web-archive-org-7310 122 19 similar similar JJ web-archive-org-7310 122 20 fashion fashion NN web-archive-org-7310 122 21 . . . web-archive-org-7310 123 1 The the DT web-archive-org-7310 123 2 document document NN web-archive-org-7310 123 3 sampling sample VBG web-archive-org-7310 123 4 module module NN web-archive-org-7310 123 5 saves save VBZ web-archive-org-7310 123 6 those those DT web-archive-org-7310 123 7 successfully successfully RB web-archive-org-7310 123 8 downloaded download VBN web-archive-org-7310 123 9 documents document NNS web-archive-org-7310 123 10 whose whose WP$ web-archive-org-7310 123 11 URLs URLs NNPS web-archive-org-7310 123 12 hash hash NN web-archive-org-7310 123 13 to to IN web-archive-org-7310 123 14 0 0 CD web-archive-org-7310 123 15 modulo modulo JJ web-archive-org-7310 123 16 1000 1000 CD web-archive-org-7310 123 17 ; ; : web-archive-org-7310 123 18 in in IN web-archive-org-7310 123 19 other other JJ web-archive-org-7310 123 20 words word NNS web-archive-org-7310 123 21 , , , web-archive-org-7310 123 22 it -PRON- PRP web-archive-org-7310 123 23 saves save VBZ web-archive-org-7310 123 24 0.1 0.1 CD web-archive-org-7310 123 25 % % NN web-archive-org-7310 123 26 of of IN web-archive-org-7310 123 27 all all DT web-archive-org-7310 123 28 downloaded download VBN web-archive-org-7310 123 29 documents document NNS web-archive-org-7310 123 30 ( ( -LRB- web-archive-org-7310 123 31 assuming assume VBG web-archive-org-7310 123 32 we -PRON- PRP web-archive-org-7310 123 33 used use VBD web-archive-org-7310 123 34 a a DT web-archive-org-7310 123 35 uniform uniform JJ web-archive-org-7310 123 36 hash hash NN web-archive-org-7310 123 37 function function NN web-archive-org-7310 123 38 ) ) -RRB- web-archive-org-7310 123 39 . . . web-archive-org-7310 124 1 The the DT web-archive-org-7310 124 2 log log NN web-archive-org-7310 124 3 contains contain VBZ web-archive-org-7310 124 4 the the DT web-archive-org-7310 124 5 URL url NN web-archive-org-7310 124 6 of of IN web-archive-org-7310 124 7 each each DT web-archive-org-7310 124 8 document document NN web-archive-org-7310 124 9 as as RB web-archive-org-7310 124 10 well well RB web-archive-org-7310 124 11 as as IN web-archive-org-7310 124 12 the the DT web-archive-org-7310 124 13 entire entire JJ web-archive-org-7310 124 14 HTTP http JJ web-archive-org-7310 124 15 response response NN web-archive-org-7310 124 16 to to IN web-archive-org-7310 124 17 each each DT web-archive-org-7310 124 18 request request NN web-archive-org-7310 124 19 , , , web-archive-org-7310 124 20 which which WDT web-archive-org-7310 124 21 includes include VBZ web-archive-org-7310 124 22 the the DT web-archive-org-7310 124 23 HTTP HTTP NNP web-archive-org-7310 124 24 header header NN web-archive-org-7310 124 25 as as RB web-archive-org-7310 124 26 well well RB web-archive-org-7310 124 27 as as IN web-archive-org-7310 124 28 the the DT web-archive-org-7310 124 29 document document NN web-archive-org-7310 124 30 . . . web-archive-org-7310 125 1 3.2 3.2 CD web-archive-org-7310 125 2       _SP web-archive-org-7310 125 3 Distilling Distilling NNP web-archive-org-7310 125 4 the the DT web-archive-org-7310 125 5 Data Data NNP web-archive-org-7310 125 6 Crawling Crawling NNP web-archive-org-7310 125 7 left leave VBD web-archive-org-7310 125 8 us -PRON- PRP web-archive-org-7310 125 9 with with IN web-archive-org-7310 125 10 44 44 CD web-archive-org-7310 125 11 very very RB web-archive-org-7310 125 12 large large JJ web-archive-org-7310 125 13 logs log NNS web-archive-org-7310 125 14 ( ( -LRB- web-archive-org-7310 125 15 produced produce VBN web-archive-org-7310 125 16 by by IN web-archive-org-7310 125 17 the the DT web-archive-org-7310 125 18 eleven eleven CD web-archive-org-7310 125 19 crawls crawl NNS web-archive-org-7310 125 20 on on IN web-archive-org-7310 125 21 the the DT web-archive-org-7310 125 22 four four CD web-archive-org-7310 125 23 crawlers crawler NNS web-archive-org-7310 125 24 ) ) -RRB- web-archive-org-7310 125 25 , , , web-archive-org-7310 125 26 each each DT web-archive-org-7310 125 27 spanning span VBG web-archive-org-7310 125 28 multiple multiple JJ web-archive-org-7310 125 29 files file NNS web-archive-org-7310 125 30 , , , web-archive-org-7310 125 31 one one CD web-archive-org-7310 125 32 per per IN web-archive-org-7310 125 33 day day NN web-archive-org-7310 125 34 . . . web-archive-org-7310 126 1 The the DT web-archive-org-7310 126 2 logs log NNS web-archive-org-7310 126 3 totaled total VBD web-archive-org-7310 126 4 about about RB web-archive-org-7310 126 5 1,200 1,200 CD web-archive-org-7310 126 6 GB GB NNP web-archive-org-7310 126 7 , , , web-archive-org-7310 126 8 whereas whereas IN web-archive-org-7310 126 9 the the DT web-archive-org-7310 126 10 sampled sample VBN web-archive-org-7310 126 11 documents document NNS web-archive-org-7310 126 12 took take VBD web-archive-org-7310 126 13 up up RP web-archive-org-7310 126 14 a a DT web-archive-org-7310 126 15 mere mere JJ web-archive-org-7310 126 16 59 59 CD web-archive-org-7310 126 17 GB GB NNP web-archive-org-7310 126 18 . . . web-archive-org-7310 127 1 As as IN web-archive-org-7310 127 2 they -PRON- PRP web-archive-org-7310 127 3 were be VBD web-archive-org-7310 127 4 , , , web-archive-org-7310 127 5 these these DT web-archive-org-7310 127 6 logs log NNS web-archive-org-7310 127 7 were be VBD web-archive-org-7310 127 8 not not RB web-archive-org-7310 127 9 suitable suitable JJ web-archive-org-7310 127 10 for for IN web-archive-org-7310 127 11 analysis analysis NN web-archive-org-7310 127 12 yet yet RB web-archive-org-7310 127 13 , , , web-archive-org-7310 127 14 because because IN web-archive-org-7310 127 15 the the DT web-archive-org-7310 127 16 URLs url NNS web-archive-org-7310 127 17 occurred occur VBD web-archive-org-7310 127 18 in in IN web-archive-org-7310 127 19 non non JJ web-archive-org-7310 127 20 - - JJ web-archive-org-7310 127 21 deterministic deterministic JJ web-archive-org-7310 127 22 order order NN web-archive-org-7310 127 23 in in IN web-archive-org-7310 127 24 each each DT web-archive-org-7310 127 25 log log NN web-archive-org-7310 127 26 . . . web-archive-org-7310 128 1 One one CD web-archive-org-7310 128 2 way way NN web-archive-org-7310 128 3 to to TO web-archive-org-7310 128 4 rectify rectify VB web-archive-org-7310 128 5 this this DT web-archive-org-7310 128 6 would would MD web-archive-org-7310 128 7 be be VB web-archive-org-7310 128 8 to to TO web-archive-org-7310 128 9 perform perform VB web-archive-org-7310 128 10 a a DT web-archive-org-7310 128 11 merge merge NN web-archive-org-7310 128 12 - - HYPH web-archive-org-7310 128 13 sort sort NN web-archive-org-7310 128 14 of of IN web-archive-org-7310 128 15 the the DT web-archive-org-7310 128 16 logs log NNS web-archive-org-7310 128 17 based base VBN web-archive-org-7310 128 18 on on IN web-archive-org-7310 128 19 their -PRON- PRP$ web-archive-org-7310 128 20 URL url NN web-archive-org-7310 128 21 , , , web-archive-org-7310 128 22 in in IN web-archive-org-7310 128 23 order order NN web-archive-org-7310 128 24 to to TO web-archive-org-7310 128 25 bring bring VB web-archive-org-7310 128 26 together together RB web-archive-org-7310 128 27 the the DT web-archive-org-7310 128 28 various various JJ web-archive-org-7310 128 29 occurrences occurrence NNS web-archive-org-7310 128 30 of of IN web-archive-org-7310 128 31 each each DT web-archive-org-7310 128 32 URL url NN web-archive-org-7310 128 33 . . . web-archive-org-7310 129 1 However however RB web-archive-org-7310 129 2 , , , web-archive-org-7310 129 3 performing perform VBG web-archive-org-7310 129 4 such such PDT web-archive-org-7310 129 5 a a DT web-archive-org-7310 129 6 merge merge NN web-archive-org-7310 129 7 - - HYPH web-archive-org-7310 129 8 sort sort NN web-archive-org-7310 129 9 on on IN web-archive-org-7310 129 10 1,200 1,200 CD web-archive-org-7310 129 11 GB GB NNP web-archive-org-7310 129 12 of of IN web-archive-org-7310 129 13 data datum NNS web-archive-org-7310 129 14 is be VBZ web-archive-org-7310 129 15 prohibitively prohibitively RB web-archive-org-7310 129 16 expensive expensive JJ web-archive-org-7310 129 17 . . . web-archive-org-7310 130 1 To to TO web-archive-org-7310 130 2 overcome overcome VB web-archive-org-7310 130 3 this this DT web-archive-org-7310 130 4 problem problem NN web-archive-org-7310 130 5 , , , web-archive-org-7310 130 6 we -PRON- PRP web-archive-org-7310 130 7 bucketized bucketize VBD web-archive-org-7310 130 8 each each DT web-archive-org-7310 130 9 of of IN web-archive-org-7310 130 10 the the DT web-archive-org-7310 130 11 logs log NNS web-archive-org-7310 130 12 , , , web-archive-org-7310 130 13 dividing divide VBG web-archive-org-7310 130 14 its -PRON- PRP$ web-archive-org-7310 130 15 contents content NNS web-archive-org-7310 130 16 over over IN web-archive-org-7310 130 17 1,000 1,000 CD web-archive-org-7310 130 18 buckets bucket NNS web-archive-org-7310 130 19 , , , web-archive-org-7310 130 20 using use VBG web-archive-org-7310 130 21 the the DT web-archive-org-7310 130 22 same same JJ web-archive-org-7310 130 23 URL url NN web-archive-org-7310 130 24 - - HYPH web-archive-org-7310 130 25 based base VBN web-archive-org-7310 130 26 hash hash NN web-archive-org-7310 130 27 function function NN web-archive-org-7310 130 28 as as IN web-archive-org-7310 130 29 in in IN web-archive-org-7310 130 30 the the DT web-archive-org-7310 130 31 document document NN web-archive-org-7310 130 32 sampling sampling NN web-archive-org-7310 130 33 module module NN web-archive-org-7310 130 34 . . . web-archive-org-7310 131 1 This this DT web-archive-org-7310 131 2 step step NN web-archive-org-7310 131 3 produced produce VBD web-archive-org-7310 131 4 44,000 44,000 CD web-archive-org-7310 131 5 buckets bucket NNS web-archive-org-7310 131 6 ( ( -LRB- web-archive-org-7310 131 7 1,000 1,000 CD web-archive-org-7310 131 8 buckets bucket NNS web-archive-org-7310 131 9 per per IN web-archive-org-7310 131 10 crawler crawler NNP web-archive-org-7310 131 11 per per IN web-archive-org-7310 131 12 crawl crawl NN web-archive-org-7310 131 13 ) ) -RRB- web-archive-org-7310 131 14 . . . web-archive-org-7310 132 1 While while IN web-archive-org-7310 132 2 this this DT web-archive-org-7310 132 3 might may MD web-archive-org-7310 132 4 appear appear VB web-archive-org-7310 132 5 to to TO web-archive-org-7310 132 6 double double VB web-archive-org-7310 132 7 the the DT web-archive-org-7310 132 8 storage storage NN web-archive-org-7310 132 9 requirement requirement NN web-archive-org-7310 132 10 , , , web-archive-org-7310 132 11 we -PRON- PRP web-archive-org-7310 132 12 could could MD web-archive-org-7310 132 13 process process VB web-archive-org-7310 132 14 an an DT web-archive-org-7310 132 15 individual individual JJ web-archive-org-7310 132 16 daily daily JJ web-archive-org-7310 132 17 log log NN web-archive-org-7310 132 18 file file NN web-archive-org-7310 132 19 , , , web-archive-org-7310 132 20 and and CC web-archive-org-7310 132 21 then then RB web-archive-org-7310 132 22 move move VB web-archive-org-7310 132 23 it -PRON- PRP web-archive-org-7310 132 24 to to IN web-archive-org-7310 132 25 near near JJ web-archive-org-7310 132 26 - - HYPH web-archive-org-7310 132 27 line line NN web-archive-org-7310 132 28 storage storage NN web-archive-org-7310 132 29 . . . web-archive-org-7310 133 1 As as IN web-archive-org-7310 133 2 a a DT web-archive-org-7310 133 3 result result NN web-archive-org-7310 133 4 of of IN web-archive-org-7310 133 5 the the DT web-archive-org-7310 133 6 bucketization bucketization NN web-archive-org-7310 133 7 , , , web-archive-org-7310 133 8 all all PDT web-archive-org-7310 133 9 the the DT web-archive-org-7310 133 10 occurrences occurrence NNS web-archive-org-7310 133 11 of of IN web-archive-org-7310 133 12 a a DT web-archive-org-7310 133 13 given give VBN web-archive-org-7310 133 14 URL url NN web-archive-org-7310 133 15 appear appear VBP web-archive-org-7310 133 16 in in IN web-archive-org-7310 133 17 corresponding correspond VBG web-archive-org-7310 133 18 buckets bucket NNS web-archive-org-7310 133 19 across across IN web-archive-org-7310 133 20 generations generation NNS web-archive-org-7310 133 21 , , , web-archive-org-7310 133 22 and and CC web-archive-org-7310 133 23 each each DT web-archive-org-7310 133 24 bucket bucket NN web-archive-org-7310 133 25 is be VBZ web-archive-org-7310 133 26 less less JJR web-archive-org-7310 133 27 than than IN web-archive-org-7310 133 28 30 30 CD web-archive-org-7310 133 29 MB MB NNS web-archive-org-7310 133 30 in in IN web-archive-org-7310 133 31 size size NN web-archive-org-7310 133 32 . . . web-archive-org-7310 134 1 This this DT web-archive-org-7310 134 2 allowed allow VBD web-archive-org-7310 134 3 us -PRON- PRP web-archive-org-7310 134 4 to to TO web-archive-org-7310 134 5 then then RB web-archive-org-7310 134 6 perform perform VB web-archive-org-7310 134 7 a a DT web-archive-org-7310 134 8 fast fast JJ web-archive-org-7310 134 9 in in IN web-archive-org-7310 134 10 - - HYPH web-archive-org-7310 134 11 memory memory NN web-archive-org-7310 134 12 sort sort NN web-archive-org-7310 134 13 of of IN web-archive-org-7310 134 14 each each DT web-archive-org-7310 134 15 bucket bucket NN web-archive-org-7310 134 16 , , , web-archive-org-7310 134 17 using use VBG web-archive-org-7310 134 18 the the DT web-archive-org-7310 134 19 URL URL NNP web-archive-org-7310 134 20 as as IN web-archive-org-7310 134 21 the the DT web-archive-org-7310 134 22 key key NN web-archive-org-7310 134 23 , , , web-archive-org-7310 134 24 and and CC web-archive-org-7310 134 25 replacing replace VBG web-archive-org-7310 134 26 the the DT web-archive-org-7310 134 27 original original JJ web-archive-org-7310 134 28 bucket bucket NN web-archive-org-7310 134 29 with with IN web-archive-org-7310 134 30 a a DT web-archive-org-7310 134 31 sorted sort VBN web-archive-org-7310 134 32 one one NN web-archive-org-7310 134 33 . . . web-archive-org-7310 135 1 At at IN web-archive-org-7310 135 2 the the DT web-archive-org-7310 135 3 end end NN web-archive-org-7310 135 4 of of IN web-archive-org-7310 135 5 this this DT web-archive-org-7310 135 6 process process NN web-archive-org-7310 135 7 , , , web-archive-org-7310 135 8 the the DT web-archive-org-7310 135 9 corresponding correspond VBG web-archive-org-7310 135 10 buckets bucket NNS web-archive-org-7310 135 11 of of IN web-archive-org-7310 135 12 each each DT web-archive-org-7310 135 13 generation generation NN web-archive-org-7310 135 14 all all DT web-archive-org-7310 135 15 contained contain VBD web-archive-org-7310 135 16 exactly exactly RB web-archive-org-7310 135 17 the the DT web-archive-org-7310 135 18 same same JJ web-archive-org-7310 135 19 set set NN web-archive-org-7310 135 20 of of IN web-archive-org-7310 135 21 URLs url NNS web-archive-org-7310 135 22 , , , web-archive-org-7310 135 23 in in IN web-archive-org-7310 135 24 sorted sorted JJ web-archive-org-7310 135 25 order order NN web-archive-org-7310 135 26 . . . web-archive-org-7310 136 1 Finally finally RB web-archive-org-7310 136 2 , , , web-archive-org-7310 136 3 we -PRON- PRP web-archive-org-7310 136 4 merged merge VBD web-archive-org-7310 136 5 the the DT web-archive-org-7310 136 6 buckets bucket NNS web-archive-org-7310 136 7 across across IN web-archive-org-7310 136 8 crawls crawl NNS web-archive-org-7310 136 9 , , , web-archive-org-7310 136 10 and and CC web-archive-org-7310 136 11 distilled distil VBD web-archive-org-7310 136 12 them -PRON- PRP web-archive-org-7310 136 13 in in IN web-archive-org-7310 136 14 the the DT web-archive-org-7310 136 15 process process NN web-archive-org-7310 136 16 . . . web-archive-org-7310 137 1 We -PRON- PRP web-archive-org-7310 137 2 did do VBD web-archive-org-7310 137 3 so so RB web-archive-org-7310 137 4 by by IN web-archive-org-7310 137 5 iterating iterate VBG web-archive-org-7310 137 6 on on IN web-archive-org-7310 137 7 all all DT web-archive-org-7310 137 8 four four CD web-archive-org-7310 137 9 machines machine NNS web-archive-org-7310 137 10 over over IN web-archive-org-7310 137 11 the the DT web-archive-org-7310 137 12 1,000 1,000 CD web-archive-org-7310 137 13 bucket bucket NN web-archive-org-7310 137 14 classes class NNS web-archive-org-7310 137 15 , , , web-archive-org-7310 137 16 reading read VBG web-archive-org-7310 137 17 a a DT web-archive-org-7310 137 18 record record NN web-archive-org-7310 137 19 at at IN web-archive-org-7310 137 20 a a DT web-archive-org-7310 137 21 time time NN web-archive-org-7310 137 22 from from IN web-archive-org-7310 137 23 each each DT web-archive-org-7310 137 24 of of IN web-archive-org-7310 137 25 the the DT web-archive-org-7310 137 26 eleven eleven CD web-archive-org-7310 137 27 corresponding corresponding JJ web-archive-org-7310 137 28 buckets bucket NNS web-archive-org-7310 137 29 , , , web-archive-org-7310 137 30 and and CC web-archive-org-7310 137 31 writing write VBG web-archive-org-7310 137 32 a a DT web-archive-org-7310 137 33 combined combine VBN web-archive-org-7310 137 34 record record NN web-archive-org-7310 137 35 to to IN web-archive-org-7310 137 36 a a DT web-archive-org-7310 137 37 distilled distilled JJ web-archive-org-7310 137 38 bucket bucket NN web-archive-org-7310 137 39 . . . web-archive-org-7310 138 1 Each each DT web-archive-org-7310 138 2 combined combined JJ web-archive-org-7310 138 3 record record NN web-archive-org-7310 138 4 contains contain VBZ web-archive-org-7310 138 5 the the DT web-archive-org-7310 138 6 following follow VBG web-archive-org-7310 138 7 information information NN web-archive-org-7310 138 8 : : : web-archive-org-7310 138 9 The the DT web-archive-org-7310 138 10 URL url NN web-archive-org-7310 138 11 . . . web-archive-org-7310 139 1 The the DT web-archive-org-7310 139 2 start start NN web-archive-org-7310 139 3 times time NNS web-archive-org-7310 139 4 of of IN web-archive-org-7310 139 5 the the DT web-archive-org-7310 139 6 eleven eleven CD web-archive-org-7310 139 7 downloads download NNS web-archive-org-7310 139 8 . . . web-archive-org-7310 140 1 The the DT web-archive-org-7310 140 2 duration duration NN web-archive-org-7310 140 3 of of IN web-archive-org-7310 140 4 each each DT web-archive-org-7310 140 5 download download NN web-archive-org-7310 140 6 . . . web-archive-org-7310 141 1 The the DT web-archive-org-7310 141 2 length length NN web-archive-org-7310 141 3 of of IN web-archive-org-7310 141 4 the the DT web-archive-org-7310 141 5 document document NN web-archive-org-7310 141 6 at at IN web-archive-org-7310 141 7 each each DT web-archive-org-7310 141 8 download download NN web-archive-org-7310 141 9 . . . web-archive-org-7310 142 1 The the DT web-archive-org-7310 142 2 number number NN web-archive-org-7310 142 3 of of IN web-archive-org-7310 142 4 non non JJ web-archive-org-7310 142 5 - - JJ web-archive-org-7310 142 6 markup markup JJ web-archive-org-7310 142 7 words word NNS web-archive-org-7310 142 8 in in IN web-archive-org-7310 142 9 the the DT web-archive-org-7310 142 10 document document NN web-archive-org-7310 142 11 at at IN web-archive-org-7310 142 12 each each DT web-archive-org-7310 142 13 download download NN web-archive-org-7310 142 14 . . . web-archive-org-7310 143 1 The the DT web-archive-org-7310 143 2 HTTP http JJ web-archive-org-7310 143 3 status status NN web-archive-org-7310 143 4 code code NN web-archive-org-7310 143 5 ( ( -LRB- web-archive-org-7310 143 6 or or CC web-archive-org-7310 143 7 error error NN web-archive-org-7310 143 8 code code NN web-archive-org-7310 143 9 ) ) -RRB- web-archive-org-7310 143 10 for for IN web-archive-org-7310 143 11 each each DT web-archive-org-7310 143 12 download download NN web-archive-org-7310 143 13 . . . web-archive-org-7310 144 1 Six six CD web-archive-org-7310 144 2 " " `` web-archive-org-7310 144 3 supershingles supershingle NNS web-archive-org-7310 144 4 " " '' web-archive-org-7310 144 5 for for IN web-archive-org-7310 144 6 each each DT web-archive-org-7310 144 7 download download NN web-archive-org-7310 144 8 . . . web-archive-org-7310 145 1 For for IN web-archive-org-7310 145 2 each each DT web-archive-org-7310 145 3 pair pair NN web-archive-org-7310 145 4 of of IN web-archive-org-7310 145 5 downloads download NNS web-archive-org-7310 145 6 , , , web-archive-org-7310 145 7 an an DT web-archive-org-7310 145 8 84-bit 84-bit CD web-archive-org-7310 145 9 vector vector NN web-archive-org-7310 145 10 indicating indicate VBG web-archive-org-7310 145 11 if if IN web-archive-org-7310 145 12 the the DT web-archive-org-7310 145 13 84 84 CD web-archive-org-7310 145 14 corresponding correspond VBG web-archive-org-7310 145 15 pre pre JJ web-archive-org-7310 145 16 - - NNS web-archive-org-7310 145 17 images image NNS web-archive-org-7310 145 18 matched match VBN web-archive-org-7310 145 19 . . . web-archive-org-7310 146 1 For for IN web-archive-org-7310 146 2 each each DT web-archive-org-7310 146 3 pair pair NN web-archive-org-7310 146 4 of of IN web-archive-org-7310 146 5 downloads download NNS web-archive-org-7310 146 6 , , , web-archive-org-7310 146 7 a a DT web-archive-org-7310 146 8 match match NN web-archive-org-7310 146 9 count count NN web-archive-org-7310 146 10 ( ( -LRB- web-archive-org-7310 146 11 see see VB web-archive-org-7310 146 12 below below RB web-archive-org-7310 146 13 ) ) -RRB- web-archive-org-7310 146 14 . . . web-archive-org-7310 147 1 Note note VB web-archive-org-7310 147 2 that that IN web-archive-org-7310 147 3 the the DT web-archive-org-7310 147 4 distilled distilled JJ web-archive-org-7310 147 5 record record NN web-archive-org-7310 147 6 does do VBZ web-archive-org-7310 147 7 not not RB web-archive-org-7310 147 8 include include VB web-archive-org-7310 147 9 the the DT web-archive-org-7310 147 10 checksums checksum NNS web-archive-org-7310 147 11 or or CC web-archive-org-7310 147 12 the the DT web-archive-org-7310 147 13 pre pre NN web-archive-org-7310 147 14 - - NNS web-archive-org-7310 147 15 images image NNS web-archive-org-7310 147 16 of of IN web-archive-org-7310 147 17 each each DT web-archive-org-7310 147 18 document document NN web-archive-org-7310 147 19 . . . web-archive-org-7310 148 1 These these DT web-archive-org-7310 148 2 values value NNS web-archive-org-7310 148 3 are be VBP web-archive-org-7310 148 4 subsumed subsume VBN web-archive-org-7310 148 5 by by IN web-archive-org-7310 148 6 the the DT web-archive-org-7310 148 7 " " `` web-archive-org-7310 148 8 supershingles supershingle NNS web-archive-org-7310 148 9 " " '' web-archive-org-7310 148 10 , , , web-archive-org-7310 148 11 the the DT web-archive-org-7310 148 12 bit bit NN web-archive-org-7310 148 13 - - HYPH web-archive-org-7310 148 14 vectors vector NNS web-archive-org-7310 148 15 , , , web-archive-org-7310 148 16 and and CC web-archive-org-7310 148 17 the the DT web-archive-org-7310 148 18 match match NN web-archive-org-7310 148 19 counts count VBZ web-archive-org-7310 148 20 . . . web-archive-org-7310 149 1 Each each DT web-archive-org-7310 149 2 of of IN web-archive-org-7310 149 3 the the DT web-archive-org-7310 149 4 six six CD web-archive-org-7310 149 5 " " `` web-archive-org-7310 149 6 supershingles supershingle NNS web-archive-org-7310 149 7 " " '' web-archive-org-7310 149 8 represents represent VBZ web-archive-org-7310 149 9 the the DT web-archive-org-7310 149 10 concatenation concatenation NN web-archive-org-7310 149 11 of of IN web-archive-org-7310 149 12 14 14 CD web-archive-org-7310 149 13 adjacent adjacent JJ web-archive-org-7310 149 14 pre pre NN web-archive-org-7310 149 15 - - NNS web-archive-org-7310 149 16 images image NNS web-archive-org-7310 149 17 . . . web-archive-org-7310 150 1 Due due IN web-archive-org-7310 150 2 to to IN web-archive-org-7310 150 3 the the DT web-archive-org-7310 150 4 independence independence NN web-archive-org-7310 150 5 of of IN web-archive-org-7310 150 6 the the DT web-archive-org-7310 150 7 one one CD web-archive-org-7310 150 8 - - HYPH web-archive-org-7310 150 9 to to IN web-archive-org-7310 150 10 - - HYPH web-archive-org-7310 150 11 one one CD web-archive-org-7310 150 12 functions function NNS web-archive-org-7310 150 13 used use VBN web-archive-org-7310 150 14 to to TO web-archive-org-7310 150 15 select select VB web-archive-org-7310 150 16 pre pre NNS web-archive-org-7310 150 17 - - NNS web-archive-org-7310 150 18 images image NNS web-archive-org-7310 150 19 , , , web-archive-org-7310 150 20 if if IN web-archive-org-7310 150 21 two two CD web-archive-org-7310 150 22 documents document NNS web-archive-org-7310 150 23 have have VBP web-archive-org-7310 150 24 similarity similarity NN web-archive-org-7310 150 25 p p NN web-archive-org-7310 150 26 , , , web-archive-org-7310 150 27 each each DT web-archive-org-7310 150 28 of of IN web-archive-org-7310 150 29 their -PRON- PRP$ web-archive-org-7310 150 30 supershingles supershingle NNS web-archive-org-7310 150 31 matches match NNS web-archive-org-7310 150 32 with with IN web-archive-org-7310 150 33 probability probability NN web-archive-org-7310 150 34 p14 p14 NN web-archive-org-7310 150 35 . . . web-archive-org-7310 151 1 For for IN web-archive-org-7310 151 2 two two CD web-archive-org-7310 151 3 documents document NNS web-archive-org-7310 151 4 that that WDT web-archive-org-7310 151 5 are be VBP web-archive-org-7310 151 6 95 95 CD web-archive-org-7310 151 7 % % NN web-archive-org-7310 151 8 similar similar JJ web-archive-org-7310 151 9 , , , web-archive-org-7310 151 10 each each DT web-archive-org-7310 151 11 supershingle supershingle NN web-archive-org-7310 151 12 matches match VBZ web-archive-org-7310 151 13 its -PRON- PRP$ web-archive-org-7310 151 14 counterpart counterpart NN web-archive-org-7310 151 15 with with IN web-archive-org-7310 151 16 probability probability NN web-archive-org-7310 151 17 49 49 CD web-archive-org-7310 151 18 % % NN web-archive-org-7310 151 19 . . . web-archive-org-7310 152 1 Given give VBN web-archive-org-7310 152 2 that that IN web-archive-org-7310 152 3 we -PRON- PRP web-archive-org-7310 152 4 retain retain VBP web-archive-org-7310 152 5 six six CD web-archive-org-7310 152 6 supershingles supershingle NNS web-archive-org-7310 152 7 , , , web-archive-org-7310 152 8 there there EX web-archive-org-7310 152 9 is be VBZ web-archive-org-7310 152 10 a a DT web-archive-org-7310 152 11 probability probability NN web-archive-org-7310 152 12 of of IN web-archive-org-7310 152 13 almost almost RB web-archive-org-7310 152 14 90 90 CD web-archive-org-7310 152 15 % % NN web-archive-org-7310 152 16 that that IN web-archive-org-7310 152 17 at at RB web-archive-org-7310 152 18 least least RBS web-archive-org-7310 152 19 two two CD web-archive-org-7310 152 20 of of IN web-archive-org-7310 152 21 the the DT web-archive-org-7310 152 22 six six CD web-archive-org-7310 152 23 supershingles supershingle NNS web-archive-org-7310 152 24 will will MD web-archive-org-7310 152 25 agree agree VB web-archive-org-7310 152 26 for for IN web-archive-org-7310 152 27 documents document NNS web-archive-org-7310 152 28 that that WDT web-archive-org-7310 152 29 are be VBP web-archive-org-7310 152 30 95 95 CD web-archive-org-7310 152 31 % % NN web-archive-org-7310 152 32 or or CC web-archive-org-7310 152 33 more more RBR web-archive-org-7310 152 34 similar similar JJ web-archive-org-7310 152 35 . . . web-archive-org-7310 153 1 Retaining retain VBG web-archive-org-7310 153 2 the the DT web-archive-org-7310 153 3 supershingles supershingle NNS web-archive-org-7310 153 4 will will MD web-archive-org-7310 153 5 be be VB web-archive-org-7310 153 6 useful useful JJ web-archive-org-7310 153 7 in in IN web-archive-org-7310 153 8 the the DT web-archive-org-7310 153 9 future future NN web-archive-org-7310 153 10 , , , web-archive-org-7310 153 11 should should MD web-archive-org-7310 153 12 we -PRON- PRP web-archive-org-7310 153 13 try try VB web-archive-org-7310 153 14 to to TO web-archive-org-7310 153 15 discover discover VB web-archive-org-7310 153 16 approximate approximate JJ web-archive-org-7310 153 17 mirrors mirror NNS web-archive-org-7310 153 18   _SP web-archive-org-7310 153 19 [ [ -LRB- web-archive-org-7310 153 20 2 2 CD web-archive-org-7310 153 21 ] ] -RRB- web-archive-org-7310 153 22 and and CC web-archive-org-7310 153 23 investigate investigate VB web-archive-org-7310 153 24 their -PRON- PRP$ web-archive-org-7310 153 25 update update NN web-archive-org-7310 153 26 frequency frequency NN web-archive-org-7310 153 27 . . . web-archive-org-7310 154 1 The the DT web-archive-org-7310 154 2 match match NN web-archive-org-7310 154 3 count count NN web-archive-org-7310 154 4 , , , web-archive-org-7310 154 5 when when WRB web-archive-org-7310 154 6 non non JJ web-archive-org-7310 154 7 - - JJ web-archive-org-7310 154 8 negative negative JJ web-archive-org-7310 154 9 , , , web-archive-org-7310 154 10 indicates indicate VBZ web-archive-org-7310 154 11 how how WRB web-archive-org-7310 154 12 many many JJ web-archive-org-7310 154 13 of of IN web-archive-org-7310 154 14 the the DT web-archive-org-7310 154 15 84 84 CD web-archive-org-7310 154 16 pre pre JJ web-archive-org-7310 154 17 - - NN web-archive-org-7310 154 18 image image NN web-archive-org-7310 154 19 pairs pair NNS web-archive-org-7310 154 20 matched match VBN web-archive-org-7310 154 21 between between IN web-archive-org-7310 154 22 the the DT web-archive-org-7310 154 23 two two CD web-archive-org-7310 154 24 documents document NNS web-archive-org-7310 154 25 . . . web-archive-org-7310 155 1 If if IN web-archive-org-7310 155 2 the the DT web-archive-org-7310 155 3 document document NN web-archive-org-7310 155 4 fingerprint fingerprint NN web-archive-org-7310 155 5 matches match NNS web-archive-org-7310 155 6 as as RB web-archive-org-7310 155 7 well well RB web-archive-org-7310 155 8 , , , web-archive-org-7310 155 9 the the DT web-archive-org-7310 155 10 match match NN web-archive-org-7310 155 11 count count NN web-archive-org-7310 155 12 is be VBZ web-archive-org-7310 155 13 set set VBN web-archive-org-7310 155 14 to to IN web-archive-org-7310 155 15 85 85 CD web-archive-org-7310 155 16 . . . web-archive-org-7310 156 1 The the DT web-archive-org-7310 156 2 match match NN web-archive-org-7310 156 3 count count NN web-archive-org-7310 156 4 is be VBZ web-archive-org-7310 156 5 negative negative JJ web-archive-org-7310 156 6 if if IN web-archive-org-7310 156 7 either either DT web-archive-org-7310 156 8 document document NN web-archive-org-7310 156 9 was be VBD web-archive-org-7310 156 10 not not RB web-archive-org-7310 156 11 downloaded download VBN web-archive-org-7310 156 12 successfully successfully RB web-archive-org-7310 156 13 or or CC web-archive-org-7310 156 14 contained contain VBN web-archive-org-7310 156 15 no no DT web-archive-org-7310 156 16 words word NNS web-archive-org-7310 156 17 at at RB web-archive-org-7310 156 18 all all RB web-archive-org-7310 156 19 ( ( -LRB- web-archive-org-7310 156 20 which which WDT web-archive-org-7310 156 21 caused cause VBD web-archive-org-7310 156 22 the the DT web-archive-org-7310 156 23 pre pre NN web-archive-org-7310 156 24 - - NNS web-archive-org-7310 156 25 images image NNS web-archive-org-7310 156 26 to to TO web-archive-org-7310 156 27 have have VB web-archive-org-7310 156 28 a a DT web-archive-org-7310 156 29 default default NN web-archive-org-7310 156 30 value value NN web-archive-org-7310 156 31 ) ) -RRB- web-archive-org-7310 156 32 ; ; : web-archive-org-7310 156 33 its -PRON- PRP$ web-archive-org-7310 156 34 value value NN web-archive-org-7310 156 35 indicates indicate VBZ web-archive-org-7310 156 36 which which WDT web-archive-org-7310 156 37 condition condition NN web-archive-org-7310 156 38 applied apply VBN web-archive-org-7310 156 39 to to IN web-archive-org-7310 156 40 which which WDT web-archive-org-7310 156 41 document document NN web-archive-org-7310 156 42 , , , web-archive-org-7310 156 43 and and CC web-archive-org-7310 156 44 whether whether IN web-archive-org-7310 156 45 the the DT web-archive-org-7310 156 46 documents document NNS web-archive-org-7310 156 47 were be VBD web-archive-org-7310 156 48 identical identical JJ web-archive-org-7310 156 49 ( ( -LRB- web-archive-org-7310 156 50 i.e. i.e. FW web-archive-org-7310 157 1 their -PRON- PRP$ web-archive-org-7310 157 2 checksums checksum NNS web-archive-org-7310 157 3 matched match VBN web-archive-org-7310 157 4 ) ) -RRB- web-archive-org-7310 157 5 . . . web-archive-org-7310 158 1 3.3 3.3 CD web-archive-org-7310 158 2       _SP web-archive-org-7310 158 3 Mining Mining NNP web-archive-org-7310 158 4 the the DT web-archive-org-7310 158 5 Data Data NNPS web-archive-org-7310 158 6 While while IN web-archive-org-7310 158 7 the the DT web-archive-org-7310 158 8 original original JJ web-archive-org-7310 158 9 logs log NNS web-archive-org-7310 158 10 were be VBD web-archive-org-7310 158 11 1,200 1,200 CD web-archive-org-7310 158 12 GB GB NNP web-archive-org-7310 158 13 in in IN web-archive-org-7310 158 14 size size NN web-archive-org-7310 158 15 , , , web-archive-org-7310 158 16 the the DT web-archive-org-7310 158 17 distilled distil VBN web-archive-org-7310 158 18 buckets bucket NNS web-archive-org-7310 158 19 consume consume VBP web-archive-org-7310 158 20 only only RB web-archive-org-7310 158 21 about about RB web-archive-org-7310 158 22 222 222 CD web-archive-org-7310 158 23 GB GB NNP web-archive-org-7310 158 24 . . . web-archive-org-7310 159 1 Even even RB web-archive-org-7310 159 2 so so RB web-archive-org-7310 159 3 , , , web-archive-org-7310 159 4 it -PRON- PRP web-archive-org-7310 159 5 still still RB web-archive-org-7310 159 6 takes take VBZ web-archive-org-7310 159 7 about about RB web-archive-org-7310 159 8 10 10 CD web-archive-org-7310 159 9 hours hour NNS web-archive-org-7310 159 10 to to TO web-archive-org-7310 159 11 read read VB web-archive-org-7310 159 12 the the DT web-archive-org-7310 159 13 distilled distilled JJ web-archive-org-7310 159 14 logs log NNS web-archive-org-7310 159 15 . . . web-archive-org-7310 160 1 In in IN web-archive-org-7310 160 2 order order NN web-archive-org-7310 160 3 to to TO web-archive-org-7310 160 4 conduct conduct VB web-archive-org-7310 160 5 the the DT web-archive-org-7310 160 6 statistical statistical JJ web-archive-org-7310 160 7 experiments experiment NNS web-archive-org-7310 160 8 we -PRON- PRP web-archive-org-7310 160 9 report report VBP web-archive-org-7310 160 10 on on IN web-archive-org-7310 160 11 below below IN web-archive-org-7310 160 12 ( ( -LRB- web-archive-org-7310 160 13 as as RB web-archive-org-7310 160 14 well well RB web-archive-org-7310 160 15 as as IN web-archive-org-7310 160 16 a a DT web-archive-org-7310 160 17 number number NN web-archive-org-7310 160 18 of of IN web-archive-org-7310 160 19 others other NNS web-archive-org-7310 160 20 which which WDT web-archive-org-7310 160 21 proved prove VBD web-archive-org-7310 160 22 less less RBR web-archive-org-7310 160 23 informative informative JJ web-archive-org-7310 160 24 ) ) -RRB- web-archive-org-7310 160 25 , , , web-archive-org-7310 160 26 we -PRON- PRP web-archive-org-7310 160 27 needed need VBD web-archive-org-7310 160 28 to to TO web-archive-org-7310 160 29 find find VB web-archive-org-7310 160 30 a a DT web-archive-org-7310 160 31 way way NN web-archive-org-7310 160 32 to to TO web-archive-org-7310 160 33 conduct conduct VB web-archive-org-7310 160 34 each each DT web-archive-org-7310 160 35 experiment experiment NN web-archive-org-7310 160 36 considerably considerably RB web-archive-org-7310 160 37 faster fast RBR web-archive-org-7310 160 38 . . . web-archive-org-7310 161 1 We -PRON- PRP web-archive-org-7310 161 2 achieve achieve VBP web-archive-org-7310 161 3 this this DT web-archive-org-7310 161 4 by by IN web-archive-org-7310 161 5 running run VBG web-archive-org-7310 161 6 several several JJ web-archive-org-7310 161 7 experiments experiment NNS web-archive-org-7310 161 8 at at IN web-archive-org-7310 161 9 once once RB web-archive-org-7310 161 10 , , , web-archive-org-7310 161 11 thereby thereby RB web-archive-org-7310 161 12 amortizing amortize VBG web-archive-org-7310 161 13 the the DT web-archive-org-7310 161 14 cost cost NN web-archive-org-7310 161 15 of of IN web-archive-org-7310 161 16 reading read VBG web-archive-org-7310 161 17 the the DT web-archive-org-7310 161 18 logs log NNS web-archive-org-7310 161 19 over over IN web-archive-org-7310 161 20 multiple multiple JJ web-archive-org-7310 161 21 experiments experiment NNS web-archive-org-7310 161 22 . . . web-archive-org-7310 162 1 Using use VBG web-archive-org-7310 162 2 this this DT web-archive-org-7310 162 3 approach approach NN web-archive-org-7310 162 4 , , , web-archive-org-7310 162 5 we -PRON- PRP web-archive-org-7310 162 6 were be VBD web-archive-org-7310 162 7 no no RB web-archive-org-7310 162 8 longer long RBR web-archive-org-7310 162 9 limited limit VBN web-archive-org-7310 162 10 by by IN web-archive-org-7310 162 11 our -PRON- PRP$ web-archive-org-7310 162 12 computer computer NN web-archive-org-7310 162 13 's 's POS web-archive-org-7310 162 14 ability ability NN web-archive-org-7310 162 15 to to TO web-archive-org-7310 162 16 read read VB web-archive-org-7310 162 17 the the DT web-archive-org-7310 162 18 logs log NNS web-archive-org-7310 162 19 , , , web-archive-org-7310 162 20 but but CC web-archive-org-7310 162 21 rather rather RB web-archive-org-7310 162 22 by by IN web-archive-org-7310 162 23 our -PRON- PRP$ web-archive-org-7310 162 24 ability ability NN web-archive-org-7310 162 25 to to TO web-archive-org-7310 162 26 decide decide VB web-archive-org-7310 162 27 a a DT web-archive-org-7310 162 28 priori priori VB web-archive-org-7310 162 29 which which WDT web-archive-org-7310 162 30 experiments experiment NNS web-archive-org-7310 162 31 would would MD web-archive-org-7310 162 32 be be VB web-archive-org-7310 162 33 interesting interesting JJ web-archive-org-7310 162 34 and and CC web-archive-org-7310 162 35 relevant relevant JJ web-archive-org-7310 162 36 . . . web-archive-org-7310 163 1 Given give VBN web-archive-org-7310 163 2 the the DT web-archive-org-7310 163 3 trial trial NN web-archive-org-7310 163 4 - - HYPH web-archive-org-7310 163 5 and and CC web-archive-org-7310 163 6 - - HYPH web-archive-org-7310 163 7 error error NN web-archive-org-7310 163 8 character character NN web-archive-org-7310 163 9 of of IN web-archive-org-7310 163 10 any any DT web-archive-org-7310 163 11 data datum NNS web-archive-org-7310 163 12 mining mining NN web-archive-org-7310 163 13 activity activity NN web-archive-org-7310 163 14 , , , web-archive-org-7310 163 15 we -PRON- PRP web-archive-org-7310 163 16 still still RB web-archive-org-7310 163 17 had have VBD web-archive-org-7310 163 18 to to TO web-archive-org-7310 163 19 do do VB web-archive-org-7310 163 20 several several JJ web-archive-org-7310 163 21 passes pass NNS web-archive-org-7310 163 22 over over IN web-archive-org-7310 163 23 the the DT web-archive-org-7310 163 24 data datum NNS web-archive-org-7310 163 25 to to TO web-archive-org-7310 163 26 construct construct VB web-archive-org-7310 163 27 all all DT web-archive-org-7310 163 28 of of IN web-archive-org-7310 163 29 the the DT web-archive-org-7310 163 30 experiments experiment NNS web-archive-org-7310 163 31 described describe VBN web-archive-org-7310 163 32 below below RB web-archive-org-7310 163 33 . . . web-archive-org-7310 164 1 We -PRON- PRP web-archive-org-7310 164 2 built build VBD web-archive-org-7310 164 3 an an DT web-archive-org-7310 164 4 analyzer analyzer NN web-archive-org-7310 164 5 harness harness NN web-archive-org-7310 164 6 that that WDT web-archive-org-7310 164 7 reads read VBZ web-archive-org-7310 164 8 through through IN web-archive-org-7310 164 9 the the DT web-archive-org-7310 164 10 distilled distil VBN web-archive-org-7310 164 11 buckets bucket NNS web-archive-org-7310 164 12 one one CD web-archive-org-7310 164 13 record record NN web-archive-org-7310 164 14 at at IN web-archive-org-7310 164 15 a a DT web-archive-org-7310 164 16 time time NN web-archive-org-7310 164 17 , , , web-archive-org-7310 164 18 expands expand VBZ web-archive-org-7310 164 19 each each DT web-archive-org-7310 164 20 record record NN web-archive-org-7310 164 21 into into IN web-archive-org-7310 164 22 an an DT web-archive-org-7310 164 23 easy easy NN web-archive-org-7310 164 24 - - HYPH web-archive-org-7310 164 25 to to TO web-archive-org-7310 164 26 - - HYPH web-archive-org-7310 164 27 use use NN web-archive-org-7310 164 28 format format NN web-archive-org-7310 164 29 , , , web-archive-org-7310 164 30 and and CC web-archive-org-7310 164 31 presents present VBZ web-archive-org-7310 164 32 it -PRON- PRP web-archive-org-7310 164 33 to to IN web-archive-org-7310 164 34 each each DT web-archive-org-7310 164 35 of of IN web-archive-org-7310 164 36 a a DT web-archive-org-7310 164 37 set set NN web-archive-org-7310 164 38 of of IN web-archive-org-7310 164 39 analyzer analyzer NNP web-archive-org-7310 164 40 modules module NNS web-archive-org-7310 164 41 . . . web-archive-org-7310 165 1 Once once IN web-archive-org-7310 165 2 all all DT web-archive-org-7310 165 3 buckets bucket NNS web-archive-org-7310 165 4 have have VBP web-archive-org-7310 165 5 been be VBN web-archive-org-7310 165 6 read read VBN web-archive-org-7310 165 7 , , , web-archive-org-7310 165 8 the the DT web-archive-org-7310 165 9 harness harness NN web-archive-org-7310 165 10 invokes invoke VBZ web-archive-org-7310 165 11 a a DT web-archive-org-7310 165 12 " " `` web-archive-org-7310 165 13 finish finish NN web-archive-org-7310 165 14 " " '' web-archive-org-7310 165 15 method method NN web-archive-org-7310 165 16 on on IN web-archive-org-7310 165 17 each each DT web-archive-org-7310 165 18 of of IN web-archive-org-7310 165 19 the the DT web-archive-org-7310 165 20 analyzers analyzer NNS web-archive-org-7310 165 21 , , , web-archive-org-7310 165 22 causing cause VBG web-archive-org-7310 165 23 them -PRON- PRP web-archive-org-7310 165 24 to to TO web-archive-org-7310 165 25 write write VB web-archive-org-7310 165 26 out out RP web-archive-org-7310 165 27 any any DT web-archive-org-7310 165 28 aggregate aggregate JJ web-archive-org-7310 165 29 statistics statistic NNS web-archive-org-7310 165 30 they -PRON- PRP web-archive-org-7310 165 31 may may MD web-archive-org-7310 165 32 have have VB web-archive-org-7310 165 33 gathered gather VBN web-archive-org-7310 165 34 . . . web-archive-org-7310 166 1 The the DT web-archive-org-7310 166 2 analyzers analyzer NNS web-archive-org-7310 166 3 as as RB web-archive-org-7310 166 4 well well RB web-archive-org-7310 166 5 as as IN web-archive-org-7310 166 6 the the DT web-archive-org-7310 166 7 harness harness NN web-archive-org-7310 166 8 are be VBP web-archive-org-7310 166 9 written write VBN web-archive-org-7310 166 10 in in IN web-archive-org-7310 166 11 Java Java NNP web-archive-org-7310 166 12 ; ; : web-archive-org-7310 166 13 the the DT web-archive-org-7310 166 14 harness harness NN web-archive-org-7310 166 15 uses use VBZ web-archive-org-7310 166 16 Java Java NNP web-archive-org-7310 166 17 's 's POS web-archive-org-7310 166 18 dynamic dynamic JJ web-archive-org-7310 166 19 class class NN web-archive-org-7310 166 20 loading loading NN web-archive-org-7310 166 21 capabilities capability NNS web-archive-org-7310 166 22 to to TO web-archive-org-7310 166 23 load load VB web-archive-org-7310 166 24 the the DT web-archive-org-7310 166 25 analyzers analyzer NNS web-archive-org-7310 166 26 at at IN web-archive-org-7310 166 27 run run NN web-archive-org-7310 166 28 - - HYPH web-archive-org-7310 166 29 time time NN web-archive-org-7310 166 30 . . . web-archive-org-7310 167 1 Examples example NNS web-archive-org-7310 167 2 of of IN web-archive-org-7310 167 3 the the DT web-archive-org-7310 167 4 analyzers analyzer NNS web-archive-org-7310 167 5 we -PRON- PRP web-archive-org-7310 167 6 wrote write VBD web-archive-org-7310 167 7 include include VBP web-archive-org-7310 167 8 : : : web-archive-org-7310 167 9 StatusCodeAnalyzer statuscodeanalyzer NN web-archive-org-7310 167 10 , , , web-archive-org-7310 167 11 which which WDT web-archive-org-7310 167 12 produces produce VBZ web-archive-org-7310 167 13 a a DT web-archive-org-7310 167 14 histogram histogram NN web-archive-org-7310 167 15 of of IN web-archive-org-7310 167 16 HTTP HTTP VBN web-archive-org-7310 167 17 status status NN web-archive-org-7310 167 18 codes code NNS web-archive-org-7310 167 19 and and CC web-archive-org-7310 167 20 TCP TCP NNP web-archive-org-7310 167 21 error error NN web-archive-org-7310 167 22 conditions condition NNS web-archive-org-7310 167 23 . . . web-archive-org-7310 168 1 FetchDurationAnalyzer FetchDurationAnalyzer NNP web-archive-org-7310 168 2 , , , web-archive-org-7310 168 3 which which WDT web-archive-org-7310 168 4 produces produce VBZ web-archive-org-7310 168 5 a a DT web-archive-org-7310 168 6 logarithmic logarithmic JJ web-archive-org-7310 168 7 histogram histogram NN web-archive-org-7310 168 8 of of IN web-archive-org-7310 168 9 document document JJ web-archive-org-7310 168 10 download download NN web-archive-org-7310 168 11 durations duration NNS web-archive-org-7310 168 12 . . . web-archive-org-7310 169 1 DocLengthAnalyzer DocLengthAnalyzer NNP web-archive-org-7310 169 2 , , , web-archive-org-7310 169 3 which which WDT web-archive-org-7310 169 4 produces produce VBZ web-archive-org-7310 169 5 a a DT web-archive-org-7310 169 6 logarithmic logarithmic JJ web-archive-org-7310 169 7 histogram histogram NN web-archive-org-7310 169 8 of of IN web-archive-org-7310 169 9 document document NN web-archive-org-7310 169 10 lengths length NNS web-archive-org-7310 169 11 . . . web-archive-org-7310 170 1 NumVersionsAnalyzer NumVersionsAnalyzer NNP web-archive-org-7310 170 2 , , , web-archive-org-7310 170 3 which which WDT web-archive-org-7310 170 4 produces produce VBZ web-archive-org-7310 170 5 a a DT web-archive-org-7310 170 6 histogram histogram NN web-archive-org-7310 170 7 of of IN web-archive-org-7310 170 8 how how WRB web-archive-org-7310 170 9 many many JJ web-archive-org-7310 170 10 versions version NNS web-archive-org-7310 170 11 of of IN web-archive-org-7310 170 12 each each DT web-archive-org-7310 170 13 URL URL NNP web-archive-org-7310 170 14 we -PRON- PRP web-archive-org-7310 170 15 managed manage VBD web-archive-org-7310 170 16 to to TO web-archive-org-7310 170 17 download download VB web-archive-org-7310 170 18 successfully successfully RB web-archive-org-7310 170 19 . . . web-archive-org-7310 171 1 TopLevelDomainAnalyzer TopLevelDomainAnalyzer NNP web-archive-org-7310 171 2 , , , web-archive-org-7310 171 3 which which WDT web-archive-org-7310 171 4 produces produce VBZ web-archive-org-7310 171 5 a a DT web-archive-org-7310 171 6 histogram histogram NN web-archive-org-7310 171 7 of of IN web-archive-org-7310 171 8 top top JJ web-archive-org-7310 171 9 - - HYPH web-archive-org-7310 171 10 level level NN web-archive-org-7310 171 11 domains domain NNS web-archive-org-7310 171 12 , , , web-archive-org-7310 171 13 and and CC web-archive-org-7310 171 14 optionally optionally RB web-archive-org-7310 171 15 counts count VBZ web-archive-org-7310 171 16 the the DT web-archive-org-7310 171 17 number number NN web-archive-org-7310 171 18 of of IN web-archive-org-7310 171 19 hosts host NNS web-archive-org-7310 171 20 in in IN web-archive-org-7310 171 21 that that DT web-archive-org-7310 171 22 domain domain NN web-archive-org-7310 171 23 . . . web-archive-org-7310 172 1 ChangeAnalyzer ChangeAnalyzer NNP web-archive-org-7310 172 2 , , , web-archive-org-7310 172 3 which which WDT web-archive-org-7310 172 4 produces produce VBZ web-archive-org-7310 172 5 a a DT web-archive-org-7310 172 6 histogram histogram NN web-archive-org-7310 172 7 of of IN web-archive-org-7310 172 8 the the DT web-archive-org-7310 172 9 number number NN web-archive-org-7310 172 10 of of IN web-archive-org-7310 172 11 unchanged unchanged JJ web-archive-org-7310 172 12 pre pre NN web-archive-org-7310 172 13 - - NNS web-archive-org-7310 172 14 images image NNS web-archive-org-7310 172 15 between between IN web-archive-org-7310 172 16 two two CD web-archive-org-7310 172 17 successive successive JJ web-archive-org-7310 172 18 successful successful JJ web-archive-org-7310 172 19 downloads download NNS web-archive-org-7310 172 20 of of IN web-archive-org-7310 172 21 a a DT web-archive-org-7310 172 22 URL url NN web-archive-org-7310 172 23 . . . web-archive-org-7310 173 1 Analyzers analyzer NNS web-archive-org-7310 173 2 can can MD web-archive-org-7310 173 3 be be VB web-archive-org-7310 173 4 nested nest VBN web-archive-org-7310 173 5 into into IN web-archive-org-7310 173 6 higher high JJR web-archive-org-7310 173 7 - - HYPH web-archive-org-7310 173 8 level level NN web-archive-org-7310 173 9 analyzers analyzer NNS web-archive-org-7310 173 10 . . . web-archive-org-7310 174 1 For for IN web-archive-org-7310 174 2 example example NN web-archive-org-7310 174 3 , , , web-archive-org-7310 174 4 it -PRON- PRP web-archive-org-7310 174 5 is be VBZ web-archive-org-7310 174 6 possible possible JJ web-archive-org-7310 174 7 to to TO web-archive-org-7310 174 8 put put VB web-archive-org-7310 174 9 a a DT web-archive-org-7310 174 10 ChangeAnalyzer ChangeAnalyzer NNP web-archive-org-7310 174 11 into into IN web-archive-org-7310 174 12 a a DT web-archive-org-7310 174 13 TopLevelDomainAnalyzer TopLevelDomainAnalyzer NNP web-archive-org-7310 174 14 , , , web-archive-org-7310 174 15 producing produce VBG web-archive-org-7310 174 16 a a DT web-archive-org-7310 174 17 list list NN web-archive-org-7310 174 18 of of IN web-archive-org-7310 174 19 document document NN web-archive-org-7310 174 20 change change NN web-archive-org-7310 174 21 histograms histogram NNS web-archive-org-7310 174 22 , , , web-archive-org-7310 174 23 one one CD web-archive-org-7310 174 24 for for IN web-archive-org-7310 174 25 each each DT web-archive-org-7310 174 26 top top JJ web-archive-org-7310 174 27 - - HYPH web-archive-org-7310 174 28 level level NN web-archive-org-7310 174 29 domain domain NN web-archive-org-7310 174 30 . . . web-archive-org-7310 175 1 This this DT web-archive-org-7310 175 2 is be VBZ web-archive-org-7310 175 3 again again RB web-archive-org-7310 175 4 implemented implement VBN web-archive-org-7310 175 5 using use VBG web-archive-org-7310 175 6 Java Java NNP web-archive-org-7310 175 7 's 's POS web-archive-org-7310 175 8 dynamic dynamic JJ web-archive-org-7310 175 9 class class NN web-archive-org-7310 175 10 loading loading NN web-archive-org-7310 175 11 machinery machinery NN web-archive-org-7310 175 12 . . . web-archive-org-7310 176 1 In in IN web-archive-org-7310 176 2 order order NN web-archive-org-7310 176 3 to to TO web-archive-org-7310 176 4 keep keep VB web-archive-org-7310 176 5 the the DT web-archive-org-7310 176 6 size size NN web-archive-org-7310 176 7 of of IN web-archive-org-7310 176 8 the the DT web-archive-org-7310 176 9 output output NN web-archive-org-7310 176 10 manageable manageable JJ web-archive-org-7310 176 11 , , , web-archive-org-7310 176 12 we -PRON- PRP web-archive-org-7310 176 13 also also RB web-archive-org-7310 176 14 provide provide VBP web-archive-org-7310 176 15 ways way NNS web-archive-org-7310 176 16 to to TO web-archive-org-7310 176 17 limit limit VB web-archive-org-7310 176 18 the the DT web-archive-org-7310 176 19 number number NN web-archive-org-7310 176 20 of of IN web-archive-org-7310 176 21 top top JJ web-archive-org-7310 176 22 - - HYPH web-archive-org-7310 176 23 level level NN web-archive-org-7310 176 24 domains domain NNS web-archive-org-7310 176 25 considered consider VBN web-archive-org-7310 176 26 ( ( -LRB- web-archive-org-7310 176 27 putting put VBG web-archive-org-7310 176 28 all all DT web-archive-org-7310 176 29 unspecified unspecified JJ web-archive-org-7310 176 30 domains domain NNS web-archive-org-7310 176 31 into into IN web-archive-org-7310 176 32 a a DT web-archive-org-7310 176 33 catch catch VB web-archive-org-7310 176 34 - - HYPH web-archive-org-7310 176 35 all all DT web-archive-org-7310 176 36 category category NN web-archive-org-7310 176 37 ) ) -RRB- web-archive-org-7310 176 38 , , , web-archive-org-7310 176 39 to to IN web-archive-org-7310 176 40 group group NN web-archive-org-7310 176 41 numbers number NNS web-archive-org-7310 176 42 of of IN web-archive-org-7310 176 43 unchanged unchanged JJ web-archive-org-7310 176 44 pre pre NN web-archive-org-7310 176 45 - - NNS web-archive-org-7310 176 46 images image NNS web-archive-org-7310 176 47 into into IN web-archive-org-7310 176 48 clusters cluster NNS web-archive-org-7310 176 49 , , , web-archive-org-7310 176 50 and and CC web-archive-org-7310 176 51 the the DT web-archive-org-7310 176 52 like like JJ web-archive-org-7310 176 53 . . . web-archive-org-7310 177 1 For for IN web-archive-org-7310 177 2 some some DT web-archive-org-7310 177 3 of of IN web-archive-org-7310 177 4 these these DT web-archive-org-7310 177 5 higher high JJR web-archive-org-7310 177 6 - - HYPH web-archive-org-7310 177 7 level level NN web-archive-org-7310 177 8 analyzers analyzer NNS web-archive-org-7310 177 9 , , , web-archive-org-7310 177 10 we -PRON- PRP web-archive-org-7310 177 11 need need VBP web-archive-org-7310 177 12 to to TO web-archive-org-7310 177 13 aggregate aggregate VB web-archive-org-7310 177 14 multiple multiple JJ web-archive-org-7310 177 15 values value NNS web-archive-org-7310 177 16 ( ( -LRB- web-archive-org-7310 177 17 for for IN web-archive-org-7310 177 18 example example NN web-archive-org-7310 177 19 , , , web-archive-org-7310 177 20 the the DT web-archive-org-7310 177 21 sizes size NNS web-archive-org-7310 177 22 of of IN web-archive-org-7310 177 23 the the DT web-archive-org-7310 177 24 eleven eleven CD web-archive-org-7310 177 25 versions version NNS web-archive-org-7310 177 26 of of IN web-archive-org-7310 177 27 a a DT web-archive-org-7310 177 28 web web NN web-archive-org-7310 177 29 page page NN web-archive-org-7310 177 30 ) ) -RRB- web-archive-org-7310 177 31 into into IN web-archive-org-7310 177 32 a a DT web-archive-org-7310 177 33 single single JJ web-archive-org-7310 177 34 value value NN web-archive-org-7310 177 35 , , , web-archive-org-7310 177 36 in in IN web-archive-org-7310 177 37 order order NN web-archive-org-7310 177 38 to to TO web-archive-org-7310 177 39 decide decide VB web-archive-org-7310 177 40 what what WDT web-archive-org-7310 177 41 lower low JJR web-archive-org-7310 177 42 - - HYPH web-archive-org-7310 177 43 level level NN web-archive-org-7310 177 44 analyzer analyzer NN web-archive-org-7310 177 45 to to TO web-archive-org-7310 177 46 invoke invoke VB web-archive-org-7310 177 47 . . . web-archive-org-7310 178 1 We -PRON- PRP web-archive-org-7310 178 2 currently currently RB web-archive-org-7310 178 3 provide provide VBP web-archive-org-7310 178 4 three three CD web-archive-org-7310 178 5 different different JJ web-archive-org-7310 178 6 ways way NNS web-archive-org-7310 178 7 to to TO web-archive-org-7310 178 8 aggregate aggregate VB web-archive-org-7310 178 9 values value NNS web-archive-org-7310 178 10 , , , web-archive-org-7310 178 11 namely namely RB web-archive-org-7310 178 12 minimum minimum JJ web-archive-org-7310 178 13 , , , web-archive-org-7310 178 14 maximum maximum JJ web-archive-org-7310 178 15 , , , web-archive-org-7310 178 16 and and CC web-archive-org-7310 178 17 average average JJ web-archive-org-7310 178 18 . . . web-archive-org-7310 179 1 We -PRON- PRP web-archive-org-7310 179 2 plan plan VBP web-archive-org-7310 179 3 to to TO web-archive-org-7310 179 4 add add VB web-archive-org-7310 179 5 support support NN web-archive-org-7310 179 6 for for IN web-archive-org-7310 179 7 mode mode NN web-archive-org-7310 179 8 , , , web-archive-org-7310 179 9 median median NNP web-archive-org-7310 179 10 , , , web-archive-org-7310 179 11 and and CC web-archive-org-7310 179 12 geometric geometric JJ web-archive-org-7310 179 13 average average NN web-archive-org-7310 179 14 . . . web-archive-org-7310 180 1 Recall recall VB web-archive-org-7310 180 2 that that IN web-archive-org-7310 180 3 during during IN web-archive-org-7310 180 4 the the DT web-archive-org-7310 180 5 data datum NNS web-archive-org-7310 180 6 collection collection NN web-archive-org-7310 180 7 phase phase NN web-archive-org-7310 180 8 , , , web-archive-org-7310 180 9 we -PRON- PRP web-archive-org-7310 180 10 saved save VBD web-archive-org-7310 180 11 the the DT web-archive-org-7310 180 12 full full JJ web-archive-org-7310 180 13 text text NN web-archive-org-7310 180 14 of of IN web-archive-org-7310 180 15 0.1 0.1 CD web-archive-org-7310 180 16 % % NN web-archive-org-7310 180 17 of of IN web-archive-org-7310 180 18 all all DT web-archive-org-7310 180 19 successfully successfully RB web-archive-org-7310 180 20 downloaded download VBN web-archive-org-7310 180 21 pages page NNS web-archive-org-7310 180 22 . . . web-archive-org-7310 181 1 We -PRON- PRP web-archive-org-7310 181 2 selected select VBD web-archive-org-7310 181 3 these these DT web-archive-org-7310 181 4 pages page NNS web-archive-org-7310 181 5 based base VBN web-archive-org-7310 181 6 on on IN web-archive-org-7310 181 7 a a DT web-archive-org-7310 181 8 hash hash NN web-archive-org-7310 181 9 of of IN web-archive-org-7310 181 10 the the DT web-archive-org-7310 181 11 URL URL NNP web-archive-org-7310 181 12 , , , web-archive-org-7310 181 13 using use VBG web-archive-org-7310 181 14 the the DT web-archive-org-7310 181 15 same same JJ web-archive-org-7310 181 16 hash hash NN web-archive-org-7310 181 17 function function NN web-archive-org-7310 181 18 as as IN web-archive-org-7310 181 19 the the DT web-archive-org-7310 181 20 bucketizer bucketizer NN web-archive-org-7310 181 21 . . . web-archive-org-7310 182 1 In in IN web-archive-org-7310 182 2 particular particular JJ web-archive-org-7310 182 3 , , , web-archive-org-7310 182 4 we -PRON- PRP web-archive-org-7310 182 5 saved save VBD web-archive-org-7310 182 6 the the DT web-archive-org-7310 182 7 full full JJ web-archive-org-7310 182 8 text text NN web-archive-org-7310 182 9 of of IN web-archive-org-7310 182 10 all all DT web-archive-org-7310 182 11 pages page NNS web-archive-org-7310 182 12 that that WDT web-archive-org-7310 182 13 went go VBD web-archive-org-7310 182 14 into into IN web-archive-org-7310 182 15 bucket bucket NN web-archive-org-7310 182 16 0 0 CD web-archive-org-7310 182 17 . . . web-archive-org-7310 183 1 This this DT web-archive-org-7310 183 2 enables enable VBZ web-archive-org-7310 183 3 us -PRON- PRP web-archive-org-7310 183 4 to to TO web-archive-org-7310 183 5 use use VB web-archive-org-7310 183 6 the the DT web-archive-org-7310 183 7 analyzer analyzer NNP web-archive-org-7310 183 8 framework framework NN web-archive-org-7310 183 9 to to TO web-archive-org-7310 183 10 detect detect VB web-archive-org-7310 183 11 interesting interesting JJ web-archive-org-7310 183 12 patterns pattern NNS web-archive-org-7310 183 13 , , , web-archive-org-7310 183 14 to to TO web-archive-org-7310 183 15 use use VB web-archive-org-7310 183 16 a a DT web-archive-org-7310 183 17 special special JJ web-archive-org-7310 183 18 analyzer analyzer NN web-archive-org-7310 183 19 ( ( -LRB- web-archive-org-7310 183 20 using use VBG web-archive-org-7310 183 21 higher high JJR web-archive-org-7310 183 22 - - HYPH web-archive-org-7310 183 23 level level NN web-archive-org-7310 183 24 analyzers analyzer NNS web-archive-org-7310 183 25 around around IN web-archive-org-7310 183 26 a a DT web-archive-org-7310 183 27 " " `` web-archive-org-7310 183 28 DumpURLAnalyzer dumpurlanalyzer NN web-archive-org-7310 183 29 " " '' web-archive-org-7310 183 30 ) ) -RRB- web-archive-org-7310 183 31 to to TO web-archive-org-7310 183 32 get get VB web-archive-org-7310 183 33 a a DT web-archive-org-7310 183 34 listing listing NN web-archive-org-7310 183 35 of of IN web-archive-org-7310 183 36 all all DT web-archive-org-7310 183 37 URLs url NNS web-archive-org-7310 183 38 in in IN web-archive-org-7310 183 39 bucket bucket NN web-archive-org-7310 183 40 0 0 CD web-archive-org-7310 183 41 that that RB web-archive-org-7310 183 42 fit fit VBP web-archive-org-7310 183 43 this this DT web-archive-org-7310 183 44 pattern pattern NN web-archive-org-7310 183 45 , , , web-archive-org-7310 183 46 and and CC web-archive-org-7310 183 47 then then RB web-archive-org-7310 183 48 to to TO web-archive-org-7310 183 49 examine examine VB web-archive-org-7310 183 50 the the DT web-archive-org-7310 183 51 full full JJ web-archive-org-7310 183 52 text text NN web-archive-org-7310 183 53 of of IN web-archive-org-7310 183 54 some some DT web-archive-org-7310 183 55 of of IN web-archive-org-7310 183 56 these these DT web-archive-org-7310 183 57 documents document NNS web-archive-org-7310 183 58 . . . web-archive-org-7310 184 1 We -PRON- PRP web-archive-org-7310 184 2 built build VBD web-archive-org-7310 184 3 some some DT web-archive-org-7310 184 4 infrastructure infrastructure NN web-archive-org-7310 184 5 to to TO web-archive-org-7310 184 6 make make VB web-archive-org-7310 184 7 this this DT web-archive-org-7310 184 8 process process NN web-archive-org-7310 184 9 easier easy JJR web-archive-org-7310 184 10 . . . web-archive-org-7310 185 1 In in IN web-archive-org-7310 185 2 particular particular JJ web-archive-org-7310 185 3 , , , web-archive-org-7310 185 4 we -PRON- PRP web-archive-org-7310 185 5 prepend prepend VBP web-archive-org-7310 185 6 each each DT web-archive-org-7310 185 7 file file NN web-archive-org-7310 185 8 of of IN web-archive-org-7310 185 9 sampled sample VBN web-archive-org-7310 185 10 documents document NNS web-archive-org-7310 185 11 with with IN web-archive-org-7310 185 12 an an DT web-archive-org-7310 185 13 index index NN web-archive-org-7310 185 14 of of IN web-archive-org-7310 185 15 URLs url NNS web-archive-org-7310 185 16 and and CC web-archive-org-7310 185 17 document document NN web-archive-org-7310 185 18 offsets offset NNS web-archive-org-7310 185 19 , , , web-archive-org-7310 185 20 which which WDT web-archive-org-7310 185 21 allows allow VBZ web-archive-org-7310 185 22 us -PRON- PRP web-archive-org-7310 185 23 to to TO web-archive-org-7310 185 24 retrieve retrieve VB web-archive-org-7310 185 25 any any DT web-archive-org-7310 185 26 page page NN web-archive-org-7310 185 27 in in IN web-archive-org-7310 185 28 constant constant JJ web-archive-org-7310 185 29 time time NN web-archive-org-7310 185 30 . . . web-archive-org-7310 186 1 Second second JJ web-archive-org-7310 186 2 , , , web-archive-org-7310 186 3 we -PRON- PRP web-archive-org-7310 186 4 implemented implement VBD web-archive-org-7310 186 5 a a DT web-archive-org-7310 186 6 web web NN web-archive-org-7310 186 7 service service NN web-archive-org-7310 186 8 that that WDT web-archive-org-7310 186 9 accepts accept VBZ web-archive-org-7310 186 10 a a DT web-archive-org-7310 186 11 URL url NN web-archive-org-7310 186 12 , , , web-archive-org-7310 186 13 a a DT web-archive-org-7310 186 14 version version NN web-archive-org-7310 186 15 number number NN web-archive-org-7310 186 16 , , , web-archive-org-7310 186 17 and and CC web-archive-org-7310 186 18 whether whether IN web-archive-org-7310 186 19 to to TO web-archive-org-7310 186 20 return return VB web-archive-org-7310 186 21 the the DT web-archive-org-7310 186 22 HTTP http JJ web-archive-org-7310 186 23 header header NN web-archive-org-7310 186 24 or or CC web-archive-org-7310 186 25 the the DT web-archive-org-7310 186 26 document document NN web-archive-org-7310 186 27 , , , web-archive-org-7310 186 28 and and CC web-archive-org-7310 186 29 returns return VBZ web-archive-org-7310 186 30 the the DT web-archive-org-7310 186 31 requested requested JJ web-archive-org-7310 186 32 item item NN web-archive-org-7310 186 33 for for IN web-archive-org-7310 186 34 display display NN web-archive-org-7310 186 35 in in IN web-archive-org-7310 186 36 the the DT web-archive-org-7310 186 37 browser browser NN web-archive-org-7310 186 38 . . . web-archive-org-7310 187 1 Third third JJ web-archive-org-7310 187 2 , , , web-archive-org-7310 187 3 we -PRON- PRP web-archive-org-7310 187 4 implemented implement VBD web-archive-org-7310 187 5 another another DT web-archive-org-7310 187 6 web web NN web-archive-org-7310 187 7 service service NN web-archive-org-7310 187 8 that that WDT web-archive-org-7310 187 9 accepts accept VBZ web-archive-org-7310 187 10 a a DT web-archive-org-7310 187 11 list list NN web-archive-org-7310 187 12 of of IN web-archive-org-7310 187 13 analyses analysis NNS web-archive-org-7310 187 14 to to TO web-archive-org-7310 187 15 run run VB web-archive-org-7310 187 16 , , , web-archive-org-7310 187 17 executes execute VBZ web-archive-org-7310 187 18 them -PRON- PRP web-archive-org-7310 187 19 on on IN web-archive-org-7310 187 20 a a DT web-archive-org-7310 187 21 subset subset NN web-archive-org-7310 187 22 of of IN web-archive-org-7310 187 23 the the DT web-archive-org-7310 187 24 distilled distil VBN web-archive-org-7310 187 25 buckets bucket NNS web-archive-org-7310 187 26 , , , web-archive-org-7310 187 27 and and CC web-archive-org-7310 187 28 returns return VBZ web-archive-org-7310 187 29 the the DT web-archive-org-7310 187 30 results result NNS web-archive-org-7310 187 31 as as IN web-archive-org-7310 187 32 a a DT web-archive-org-7310 187 33 web web NN web-archive-org-7310 187 34 page page NN web-archive-org-7310 187 35 . . . web-archive-org-7310 188 1 We -PRON- PRP web-archive-org-7310 188 2 hope hope VBP web-archive-org-7310 188 3 to to TO web-archive-org-7310 188 4 make make VB web-archive-org-7310 188 5 these these DT web-archive-org-7310 188 6 services service NNS web-archive-org-7310 188 7 , , , web-archive-org-7310 188 8 as as RB web-archive-org-7310 188 9 well well RB web-archive-org-7310 188 10 as as IN web-archive-org-7310 188 11 some some DT web-archive-org-7310 188 12 of of IN web-archive-org-7310 188 13 the the DT web-archive-org-7310 188 14 distilled distilled JJ web-archive-org-7310 188 15 buckets bucket NNS web-archive-org-7310 188 16 , , , web-archive-org-7310 188 17 available available JJ web-archive-org-7310 188 18 to to IN web-archive-org-7310 188 19 the the DT web-archive-org-7310 188 20 research research NN web-archive-org-7310 188 21 community community NN web-archive-org-7310 188 22 . . . web-archive-org-7310 189 1 4 4 LS web-archive-org-7310 189 2 . . . web-archive-org-7310 190 1 RESULTS RESULTS NNP web-archive-org-7310 190 2 The the DT web-archive-org-7310 190 3 results result NNS web-archive-org-7310 190 4 presented present VBN web-archive-org-7310 190 5 in in IN web-archive-org-7310 190 6 this this DT web-archive-org-7310 190 7 section section NN web-archive-org-7310 190 8 are be VBP web-archive-org-7310 190 9 derived derive VBN web-archive-org-7310 190 10 from from IN web-archive-org-7310 190 11 analyzing analyze VBG web-archive-org-7310 190 12 the the DT web-archive-org-7310 190 13 151 151 CD web-archive-org-7310 190 14 million million CD web-archive-org-7310 190 15 distilled distilled JJ web-archive-org-7310 190 16 records record NNS web-archive-org-7310 190 17 in in IN web-archive-org-7310 190 18 our -PRON- PRP$ web-archive-org-7310 190 19 collection collection NN web-archive-org-7310 190 20 , , , web-archive-org-7310 190 21 using use VBG web-archive-org-7310 190 22 the the DT web-archive-org-7310 190 23 analyzer analyzer NNP web-archive-org-7310 190 24 harness harness NNP web-archive-org-7310 190 25 and and CC web-archive-org-7310 190 26 many many JJ web-archive-org-7310 190 27 of of IN web-archive-org-7310 190 28 the the DT web-archive-org-7310 190 29 analyzers analyzer NNS web-archive-org-7310 190 30 described describe VBN web-archive-org-7310 190 31 above above RB web-archive-org-7310 190 32 . . . web-archive-org-7310 191 1 Figure figure NN web-archive-org-7310 191 2   _SP web-archive-org-7310 191 3 2 2 CD web-archive-org-7310 191 4 shows show VBZ web-archive-org-7310 191 5 a a DT web-archive-org-7310 191 6 histogram histogram NN web-archive-org-7310 191 7 of of IN web-archive-org-7310 191 8 document document NN web-archive-org-7310 191 9 length length NN web-archive-org-7310 191 10 , , , web-archive-org-7310 191 11 for for IN web-archive-org-7310 191 12 all all DT web-archive-org-7310 191 13 of of IN web-archive-org-7310 191 14 the the DT web-archive-org-7310 191 15 1,482,416,213 1,482,416,213 CD web-archive-org-7310 191 16 documents document NNS web-archive-org-7310 191 17 that that WDT web-archive-org-7310 191 18 were be VBD web-archive-org-7310 191 19 successfully successfully RB web-archive-org-7310 191 20 downloaded download VBN web-archive-org-7310 191 21 ( ( -LRB- web-archive-org-7310 191 22 i.e. i.e. FW web-archive-org-7310 192 1 that that DT web-archive-org-7310 192 2 had have VBD web-archive-org-7310 192 3 an an DT web-archive-org-7310 192 4 HTTP http JJ web-archive-org-7310 192 5 status status NN web-archive-org-7310 192 6 code code NN web-archive-org-7310 192 7 of of IN web-archive-org-7310 192 8 200 200 CD web-archive-org-7310 192 9 ) ) -RRB- web-archive-org-7310 192 10 , , , web-archive-org-7310 192 11 as as RB web-archive-org-7310 192 12 well well RB web-archive-org-7310 192 13 as as IN web-archive-org-7310 192 14 broken break VBN web-archive-org-7310 192 15 out out RP web-archive-org-7310 192 16 by by IN web-archive-org-7310 192 17 a a DT web-archive-org-7310 192 18 few few JJ web-archive-org-7310 192 19 selected select VBN web-archive-org-7310 192 20 top top JJ web-archive-org-7310 192 21 - - HYPH web-archive-org-7310 192 22 level level NN web-archive-org-7310 192 23 domains domain NNS web-archive-org-7310 192 24 . . . web-archive-org-7310 193 1 The the DT web-archive-org-7310 193 2 x x NN web-archive-org-7310 193 3 - - : web-archive-org-7310 193 4 axis axis RB web-archive-org-7310 193 5 denotes denote VBZ web-archive-org-7310 193 6 the the DT web-archive-org-7310 193 7 document document NN web-archive-org-7310 193 8 size size NN web-archive-org-7310 193 9 ; ; : web-archive-org-7310 193 10 a a DT web-archive-org-7310 193 11 value value NN web-archive-org-7310 193 12 of of IN web-archive-org-7310 193 13 n n JJ web-archive-org-7310 193 14 means mean NNS web-archive-org-7310 193 15 that that IN web-archive-org-7310 193 16 the the DT web-archive-org-7310 193 17 size size NN web-archive-org-7310 193 18 of of IN web-archive-org-7310 193 19 the the DT web-archive-org-7310 193 20 document document NN web-archive-org-7310 193 21 was be VBD web-archive-org-7310 193 22 below below IN web-archive-org-7310 193 23 2n 2n CD web-archive-org-7310 193 24 bytes byte NNS web-archive-org-7310 193 25 , , , web-archive-org-7310 193 26 but but CC web-archive-org-7310 193 27 not not RB web-archive-org-7310 193 28 below below IN web-archive-org-7310 193 29 2n-1 2n-1 CD web-archive-org-7310 193 30 bytes byte NNS web-archive-org-7310 193 31 ( ( -LRB- web-archive-org-7310 193 32 a a DT web-archive-org-7310 193 33 value value NN web-archive-org-7310 193 34 of of IN web-archive-org-7310 193 35 0 0 CD web-archive-org-7310 193 36 indicates indicate VBZ web-archive-org-7310 193 37 that that IN web-archive-org-7310 193 38 the the DT web-archive-org-7310 193 39 document document NN web-archive-org-7310 193 40 had have VBD web-archive-org-7310 193 41 length length NN web-archive-org-7310 193 42 0 0 CD web-archive-org-7310 193 43 ) ) -RRB- web-archive-org-7310 193 44 . . . web-archive-org-7310 194 1 Figure figure NN web-archive-org-7310 194 2 2 2 CD web-archive-org-7310 194 3 : : : web-archive-org-7310 194 4 Distribution distribution NN web-archive-org-7310 194 5 of of IN web-archive-org-7310 194 6 documents document NNS web-archive-org-7310 194 7 lengths length NNS web-archive-org-7310 194 8 overall overall RB web-archive-org-7310 194 9 and and CC web-archive-org-7310 194 10 for for IN web-archive-org-7310 194 11 selected select VBN web-archive-org-7310 194 12 top top JJ web-archive-org-7310 194 13 - - HYPH web-archive-org-7310 194 14 level level NN web-archive-org-7310 194 15 domains domain NNS web-archive-org-7310 194 16 . . . web-archive-org-7310 195 1 The the DT web-archive-org-7310 195 2 distribution distribution NN web-archive-org-7310 195 3 we -PRON- PRP web-archive-org-7310 195 4 observed observe VBD web-archive-org-7310 195 5 centers center NNS web-archive-org-7310 195 6 at at IN web-archive-org-7310 195 7 14 14 CD web-archive-org-7310 195 8 with with IN web-archive-org-7310 195 9 standard standard JJ web-archive-org-7310 195 10 deviation deviation NN web-archive-org-7310 195 11 1 1 CD web-archive-org-7310 195 12 ; ; : web-archive-org-7310 195 13 66.3 66.3 CD web-archive-org-7310 195 14 % % NN web-archive-org-7310 195 15 of of IN web-archive-org-7310 195 16 all all DT web-archive-org-7310 195 17 observed observe VBN web-archive-org-7310 195 18 HTML html NN web-archive-org-7310 195 19 pages page NNS web-archive-org-7310 195 20 are be VBP web-archive-org-7310 195 21 between between IN web-archive-org-7310 195 22 4 4 CD web-archive-org-7310 195 23 KB KB NNP web-archive-org-7310 195 24 and and CC web-archive-org-7310 195 25 32 32 CD web-archive-org-7310 195 26 KB kb NN web-archive-org-7310 195 27 in in IN web-archive-org-7310 195 28 length length NN web-archive-org-7310 195 29 . . . web-archive-org-7310 196 1 Looking look VBG web-archive-org-7310 196 2 at at IN web-archive-org-7310 196 3 selected select VBN web-archive-org-7310 196 4 top top JJ web-archive-org-7310 196 5 - - HYPH web-archive-org-7310 196 6 level level NN web-archive-org-7310 196 7 domains domain NNS web-archive-org-7310 196 8 , , , web-archive-org-7310 196 9 pages page NNS web-archive-org-7310 196 10 in in IN web-archive-org-7310 196 11 .com .com NNP web-archive-org-7310 196 12 , , , web-archive-org-7310 196 13 which which WDT web-archive-org-7310 196 14 represent represent VBP web-archive-org-7310 196 15 52.5 52.5 CD web-archive-org-7310 196 16 % % NN web-archive-org-7310 196 17 of of IN web-archive-org-7310 196 18 all all DT web-archive-org-7310 196 19 observed observe VBN web-archive-org-7310 196 20 pages page NNS web-archive-org-7310 196 21 , , , web-archive-org-7310 196 22 largely largely RB web-archive-org-7310 196 23 reflect reflect VB web-archive-org-7310 196 24 the the DT web-archive-org-7310 196 25 overall overall JJ web-archive-org-7310 196 26 distribution distribution NN web-archive-org-7310 196 27 , , , web-archive-org-7310 196 28 but but CC web-archive-org-7310 196 29 are be VBP web-archive-org-7310 196 30 biased biased JJ web-archive-org-7310 196 31 slightly slightly RB web-archive-org-7310 196 32 higher high JJR web-archive-org-7310 196 33 . . . web-archive-org-7310 197 1 Pages page NNS web-archive-org-7310 197 2 in in IN web-archive-org-7310 197 3 .org .org . web-archive-org-7310 197 4 and and CC web-archive-org-7310 197 5 .gov .gov CD web-archive-org-7310 197 6 , , , web-archive-org-7310 197 7 which which WDT web-archive-org-7310 197 8 account account VBP web-archive-org-7310 197 9 for for IN web-archive-org-7310 197 10 8.0 8.0 CD web-archive-org-7310 197 11 % % NN web-archive-org-7310 197 12 and and CC web-archive-org-7310 197 13 1.1 1.1 CD web-archive-org-7310 197 14 % % NN web-archive-org-7310 197 15 of of IN web-archive-org-7310 197 16 all all DT web-archive-org-7310 197 17 observed observe VBN web-archive-org-7310 197 18 pages page NNS web-archive-org-7310 197 19 , , , web-archive-org-7310 197 20 respectively respectively RB web-archive-org-7310 197 21 , , , web-archive-org-7310 197 22 are be VBP web-archive-org-7310 197 23 similar similar JJ web-archive-org-7310 197 24 to to IN web-archive-org-7310 197 25 the the DT web-archive-org-7310 197 26 overall overall JJ web-archive-org-7310 197 27 distribution distribution NN web-archive-org-7310 197 28 , , , web-archive-org-7310 197 29 but but CC web-archive-org-7310 197 30 are be VBP web-archive-org-7310 197 31 biased biased JJ web-archive-org-7310 197 32 slightly slightly RB web-archive-org-7310 197 33 lower low JJR web-archive-org-7310 197 34 . . . web-archive-org-7310 198 1 Pages page NNS web-archive-org-7310 198 2 in in IN web-archive-org-7310 198 3 .edu .edu NFP web-archive-org-7310 198 4 tend tend VB web-archive-org-7310 198 5 to to TO web-archive-org-7310 198 6 be be VB web-archive-org-7310 198 7 smaller small JJR web-archive-org-7310 198 8 , , , web-archive-org-7310 198 9 with with IN web-archive-org-7310 198 10 64.9 64.9 CD web-archive-org-7310 198 11 % % NN web-archive-org-7310 198 12 of of IN web-archive-org-7310 198 13 the the DT web-archive-org-7310 198 14 pages page NNS web-archive-org-7310 198 15 being be VBG web-archive-org-7310 198 16 between between IN web-archive-org-7310 198 17 2 2 CD web-archive-org-7310 198 18 KB KB NNP web-archive-org-7310 198 19 and and CC web-archive-org-7310 198 20 16 16 CD web-archive-org-7310 198 21 KB KB NNP web-archive-org-7310 198 22 . . . web-archive-org-7310 199 1 Figure figure NN web-archive-org-7310 199 2   _SP web-archive-org-7310 199 3 3 3 CD web-archive-org-7310 199 4 is be VBZ web-archive-org-7310 199 5 similar similar JJ web-archive-org-7310 199 6 to to IN web-archive-org-7310 199 7 Figure figure NN web-archive-org-7310 199 8   _SP web-archive-org-7310 199 9 2 2 CD web-archive-org-7310 199 10 , , , web-archive-org-7310 199 11 but but CC web-archive-org-7310 199 12 shows show VBZ web-archive-org-7310 199 13 a a DT web-archive-org-7310 199 14 histogram histogram NN web-archive-org-7310 199 15 of of IN web-archive-org-7310 199 16 the the DT web-archive-org-7310 199 17 number number NN web-archive-org-7310 199 18 of of IN web-archive-org-7310 199 19 words word NNS web-archive-org-7310 199 20 per per IN web-archive-org-7310 199 21 document document NN web-archive-org-7310 199 22 , , , web-archive-org-7310 199 23 instead instead RB web-archive-org-7310 199 24 of of IN web-archive-org-7310 199 25 the the DT web-archive-org-7310 199 26 number number NN web-archive-org-7310 199 27 of of IN web-archive-org-7310 199 28 bytes byte NNS web-archive-org-7310 199 29 . . . web-archive-org-7310 200 1 Note note VB web-archive-org-7310 200 2 that that IN web-archive-org-7310 200 3 the the DT web-archive-org-7310 200 4 distribution distribution NN web-archive-org-7310 200 5 of of IN web-archive-org-7310 200 6 the the DT web-archive-org-7310 200 7 .edu .edu NFP web-archive-org-7310 200 8 domain domain NN web-archive-org-7310 200 9 is be VBZ web-archive-org-7310 200 10 closer close JJR web-archive-org-7310 200 11 to to IN web-archive-org-7310 200 12 the the DT web-archive-org-7310 200 13 overall overall JJ web-archive-org-7310 200 14 distribution distribution NN web-archive-org-7310 200 15 when when WRB web-archive-org-7310 200 16 it -PRON- PRP web-archive-org-7310 200 17 comes come VBZ web-archive-org-7310 200 18 to to IN web-archive-org-7310 200 19 words word NNS web-archive-org-7310 200 20 , , , web-archive-org-7310 200 21 suggesting suggest VBG web-archive-org-7310 200 22 that that IN web-archive-org-7310 200 23 pages page NNS web-archive-org-7310 200 24 in in IN web-archive-org-7310 200 25 .edu .edu : web-archive-org-7310 200 26 either either CC web-archive-org-7310 200 27 have have VBP web-archive-org-7310 200 28 shorter short JJR web-archive-org-7310 200 29 words word NNS web-archive-org-7310 200 30 or or CC web-archive-org-7310 200 31 less less JJR web-archive-org-7310 200 32 HTML html NN web-archive-org-7310 200 33 markup markup NN web-archive-org-7310 200 34 . . . web-archive-org-7310 201 1 Figure figure NN web-archive-org-7310 201 2 3 3 CD web-archive-org-7310 201 3 : : : web-archive-org-7310 201 4 Distribution distribution NN web-archive-org-7310 201 5 of of IN web-archive-org-7310 201 6 words word NNS web-archive-org-7310 201 7 per per IN web-archive-org-7310 201 8 documents document NNS web-archive-org-7310 201 9 overall overall JJ web-archive-org-7310 201 10 and and CC web-archive-org-7310 201 11 for for IN web-archive-org-7310 201 12 selected select VBN web-archive-org-7310 201 13 top top JJ web-archive-org-7310 201 14 - - HYPH web-archive-org-7310 201 15 level level NN web-archive-org-7310 201 16 domains domain NNS web-archive-org-7310 201 17 . . . web-archive-org-7310 202 1 Figures figure NNS web-archive-org-7310 202 2   _SP web-archive-org-7310 202 3 4 4 CD web-archive-org-7310 202 4 , , , web-archive-org-7310 202 5 5 5 CD web-archive-org-7310 202 6 , , , web-archive-org-7310 202 7 and and CC web-archive-org-7310 202 8 6 6 CD web-archive-org-7310 202 9 attempt attempt NN web-archive-org-7310 202 10 to to TO web-archive-org-7310 202 11 capture capture VB web-archive-org-7310 202 12 different different JJ web-archive-org-7310 202 13 aspects aspect NNS web-archive-org-7310 202 14 of of IN web-archive-org-7310 202 15 the the DT web-archive-org-7310 202 16 permanence permanence NN web-archive-org-7310 202 17 of of IN web-archive-org-7310 202 18 web web NN web-archive-org-7310 202 19 pages page NNS web-archive-org-7310 202 20 . . . web-archive-org-7310 203 1 Figure figure NN web-archive-org-7310 203 2   _SP web-archive-org-7310 203 3 4 4 CD web-archive-org-7310 203 4 shows show NNS web-archive-org-7310 203 5 for for IN web-archive-org-7310 203 6 each each DT web-archive-org-7310 203 7 crawl crawl JJ web-archive-org-7310 203 8 generation generation NN web-archive-org-7310 203 9 the the DT web-archive-org-7310 203 10 percentage percentage NN web-archive-org-7310 203 11 of of IN web-archive-org-7310 203 12 page page NN web-archive-org-7310 203 13 retrievals retrieval NNS web-archive-org-7310 203 14 resulting result VBG web-archive-org-7310 203 15 in in IN web-archive-org-7310 203 16 different different JJ web-archive-org-7310 203 17 categories category NNS web-archive-org-7310 203 18 of of IN web-archive-org-7310 203 19 status status NN web-archive-org-7310 203 20 codes code NNS web-archive-org-7310 203 21 . . . web-archive-org-7310 204 1 The the DT web-archive-org-7310 204 2 200 200 CD web-archive-org-7310 204 3 category category NN web-archive-org-7310 204 4 ( ( -LRB- web-archive-org-7310 204 5 corresponding correspond VBG web-archive-org-7310 204 6 to to IN web-archive-org-7310 204 7 the the DT web-archive-org-7310 204 8 HTTP http JJ web-archive-org-7310 204 9 status status NN web-archive-org-7310 204 10 code code NN web-archive-org-7310 204 11 " " `` web-archive-org-7310 204 12 200 200 CD web-archive-org-7310 204 13 OK OK NNP web-archive-org-7310 204 14 " " '' web-archive-org-7310 204 15 ) ) -RRB- web-archive-org-7310 204 16 shows show VBZ web-archive-org-7310 204 17 pages page NNS web-archive-org-7310 204 18 that that WDT web-archive-org-7310 204 19 were be VBD web-archive-org-7310 204 20 successfully successfully RB web-archive-org-7310 204 21 downloaded download VBN web-archive-org-7310 204 22 . . . web-archive-org-7310 205 1 Note note VB web-archive-org-7310 205 2 that that IN web-archive-org-7310 205 3 the the DT web-archive-org-7310 205 4 y y NN web-archive-org-7310 205 5 - - HYPH web-archive-org-7310 205 6 axis axis NNP web-archive-org-7310 205 7 starts start VBZ web-archive-org-7310 205 8 at at IN web-archive-org-7310 205 9 85 85 CD web-archive-org-7310 205 10 % % NN web-archive-org-7310 205 11 ; ; : web-archive-org-7310 205 12 in in IN web-archive-org-7310 205 13 all all DT web-archive-org-7310 205 14 generations generation NNS web-archive-org-7310 205 15 , , , web-archive-org-7310 205 16 over over IN web-archive-org-7310 205 17 85 85 CD web-archive-org-7310 205 18 % % NN web-archive-org-7310 205 19 of of IN web-archive-org-7310 205 20 the the DT web-archive-org-7310 205 21 retrievals retrieval NNS web-archive-org-7310 205 22 were be VBD web-archive-org-7310 205 23 successful successful JJ web-archive-org-7310 205 24 . . . web-archive-org-7310 206 1 The the DT web-archive-org-7310 206 2 3xx 3xx JJ web-archive-org-7310 206 3 category category NN web-archive-org-7310 206 4 contains contain VBZ web-archive-org-7310 206 5 all all PDT web-archive-org-7310 206 6 those those DT web-archive-org-7310 206 7 pages page NNS web-archive-org-7310 206 8 that that WDT web-archive-org-7310 206 9 return return VBP web-archive-org-7310 206 10 an an DT web-archive-org-7310 206 11 HTTP http JJ web-archive-org-7310 206 12 status status NN web-archive-org-7310 206 13 code code NN web-archive-org-7310 206 14 indicating indicate VBG web-archive-org-7310 206 15 that that IN web-archive-org-7310 206 16 the the DT web-archive-org-7310 206 17 page page NN web-archive-org-7310 206 18 has have VBZ web-archive-org-7310 206 19 moved move VBN web-archive-org-7310 206 20 . . . web-archive-org-7310 207 1 Since since IN web-archive-org-7310 207 2 all all DT web-archive-org-7310 207 3 URLs url NNS web-archive-org-7310 207 4 in in IN web-archive-org-7310 207 5 our -PRON- PRP$ web-archive-org-7310 207 6 set set NN web-archive-org-7310 207 7 produced produce VBD web-archive-org-7310 207 8 a a DT web-archive-org-7310 207 9 status status NN web-archive-org-7310 207 10 code code NN web-archive-org-7310 207 11 of of IN web-archive-org-7310 207 12 200 200 CD web-archive-org-7310 207 13 during during IN web-archive-org-7310 207 14 crawl crawl NNP web-archive-org-7310 207 15 1 1 CD web-archive-org-7310 207 16 , , , web-archive-org-7310 207 17 the the DT web-archive-org-7310 207 18 page page NN web-archive-org-7310 207 19 has have VBZ web-archive-org-7310 207 20 moved move VBN web-archive-org-7310 207 21 since since RB web-archive-org-7310 207 22 . . . web-archive-org-7310 208 1 The the DT web-archive-org-7310 208 2 4xx 4xx JJ web-archive-org-7310 208 3 category category NN web-archive-org-7310 208 4 contains contain VBZ web-archive-org-7310 208 5 all all DT web-archive-org-7310 208 6 client client NN web-archive-org-7310 208 7 errors error NNS web-archive-org-7310 208 8 . . . web-archive-org-7310 209 1 The the DT web-archive-org-7310 209 2 most most RBS web-archive-org-7310 209 3 common common JJ web-archive-org-7310 209 4 one one NN web-archive-org-7310 209 5 is be VBZ web-archive-org-7310 209 6 404 404 CD web-archive-org-7310 209 7 ( ( -LRB- web-archive-org-7310 209 8 " " `` web-archive-org-7310 209 9 Not not RB web-archive-org-7310 209 10 Found find VBN web-archive-org-7310 209 11 " " '' web-archive-org-7310 209 12 ) ) -RRB- web-archive-org-7310 209 13 , , , web-archive-org-7310 209 14 indicating indicate VBG web-archive-org-7310 209 15 that that IN web-archive-org-7310 209 16 the the DT web-archive-org-7310 209 17 page page NN web-archive-org-7310 209 18 has have VBZ web-archive-org-7310 209 19 disappeared disappear VBN web-archive-org-7310 209 20 without without IN web-archive-org-7310 209 21 leaving leave VBG web-archive-org-7310 209 22 a a DT web-archive-org-7310 209 23 forwarding forwarding JJ web-archive-org-7310 209 24 address address NN web-archive-org-7310 209 25 , , , web-archive-org-7310 209 26 distantly distantly RB web-archive-org-7310 209 27 followed follow VBN web-archive-org-7310 209 28 by by IN web-archive-org-7310 209 29 403 403 CD web-archive-org-7310 209 30 ( ( -LRB- web-archive-org-7310 209 31 " " `` web-archive-org-7310 209 32 Forbidden Forbidden NNP web-archive-org-7310 209 33 " " '' web-archive-org-7310 209 34 ) ) -RRB- web-archive-org-7310 209 35 . . . web-archive-org-7310 210 1 The the DT web-archive-org-7310 210 2 other other JJ web-archive-org-7310 210 3 category category NN web-archive-org-7310 210 4 contains contain VBZ web-archive-org-7310 210 5 all all DT web-archive-org-7310 210 6 pages page NNS web-archive-org-7310 210 7 for for IN web-archive-org-7310 210 8 which which WDT web-archive-org-7310 210 9 the the DT web-archive-org-7310 210 10 web web NN web-archive-org-7310 210 11 server server NN web-archive-org-7310 210 12 returned return VBD web-archive-org-7310 210 13 a a DT web-archive-org-7310 210 14 status status NN web-archive-org-7310 210 15 code code NN web-archive-org-7310 210 16 not not RB web-archive-org-7310 210 17 listed list VBN web-archive-org-7310 210 18 above above RB web-archive-org-7310 210 19 . . . web-archive-org-7310 211 1 The the DT web-archive-org-7310 211 2 various various JJ web-archive-org-7310 211 3 5xx 5xx JJ web-archive-org-7310 211 4 return return NN web-archive-org-7310 211 5 codes code NNS web-archive-org-7310 211 6 dominated dominate VBD web-archive-org-7310 211 7 ; ; : web-archive-org-7310 211 8 we -PRON- PRP web-archive-org-7310 211 9 also also RB web-archive-org-7310 211 10 found find VBD web-archive-org-7310 211 11 many many JJ web-archive-org-7310 211 12 web web NN web-archive-org-7310 211 13 servers server NNS web-archive-org-7310 211 14 returning return VBG web-archive-org-7310 211 15 status status NN web-archive-org-7310 211 16 codes code NNS web-archive-org-7310 211 17 not not RB web-archive-org-7310 211 18 listed list VBN web-archive-org-7310 211 19 in in IN web-archive-org-7310 211 20 the the DT web-archive-org-7310 211 21 HTTP http JJ web-archive-org-7310 211 22 RFC rfc NN web-archive-org-7310 211 23 . . . web-archive-org-7310 212 1 The the DT web-archive-org-7310 212 2 network network NN web-archive-org-7310 212 3 category category NN web-archive-org-7310 212 4 contains contain VBZ web-archive-org-7310 212 5 all all DT web-archive-org-7310 212 6 retrieval retrieval NN web-archive-org-7310 212 7 attempts attempt NNS web-archive-org-7310 212 8 that that WDT web-archive-org-7310 212 9 failed fail VBD web-archive-org-7310 212 10 due due IN web-archive-org-7310 212 11 to to IN web-archive-org-7310 212 12 a a DT web-archive-org-7310 212 13 network network NN web-archive-org-7310 212 14 - - HYPH web-archive-org-7310 212 15 related relate VBN web-archive-org-7310 212 16 error error NN web-archive-org-7310 212 17 , , , web-archive-org-7310 212 18 such such JJ web-archive-org-7310 212 19 as as IN web-archive-org-7310 212 20 DNS DNS NNP web-archive-org-7310 212 21 lookup lookup NN web-archive-org-7310 212 22 failure failure NN web-archive-org-7310 212 23 , , , web-archive-org-7310 212 24 refused refuse VBD web-archive-org-7310 212 25 connections connection NNS web-archive-org-7310 212 26 , , , web-archive-org-7310 212 27 TCP TCP NNP web-archive-org-7310 212 28 timeouts timeout NNS web-archive-org-7310 212 29 , , , web-archive-org-7310 212 30 and and CC web-archive-org-7310 212 31 the the DT web-archive-org-7310 212 32 like like JJ web-archive-org-7310 212 33 ( ( -LRB- web-archive-org-7310 212 34 note note VB web-archive-org-7310 212 35 that that IN web-archive-org-7310 212 36 Mercator Mercator NNP web-archive-org-7310 212 37 makes make VBZ web-archive-org-7310 212 38 five five CD web-archive-org-7310 212 39 attempts attempt NNS web-archive-org-7310 212 40 to to TO web-archive-org-7310 212 41 resolve resolve VB web-archive-org-7310 212 42 a a DT web-archive-org-7310 212 43 domain domain NN web-archive-org-7310 212 44 name name NN web-archive-org-7310 212 45 , , , web-archive-org-7310 212 46 and and CC web-archive-org-7310 212 47 three three CD web-archive-org-7310 212 48 attempts attempt NNS web-archive-org-7310 212 49 at at IN web-archive-org-7310 212 50 retrievals retrieval NNS web-archive-org-7310 212 51 that that WDT web-archive-org-7310 212 52 fail fail VBP web-archive-org-7310 212 53 due due IN web-archive-org-7310 212 54 to to IN web-archive-org-7310 212 55 TCP TCP NNP web-archive-org-7310 212 56 errors error NNS web-archive-org-7310 212 57 ) ) -RRB- web-archive-org-7310 212 58 . . . web-archive-org-7310 213 1 The the DT web-archive-org-7310 213 2 RobotExcl RobotExcl NNP web-archive-org-7310 213 3 category category NN web-archive-org-7310 213 4 contains contain VBZ web-archive-org-7310 213 5 all all PDT web-archive-org-7310 213 6 those those DT web-archive-org-7310 213 7 pages page NNS web-archive-org-7310 213 8 that that WDT web-archive-org-7310 213 9 were be VBD web-archive-org-7310 213 10 blocked block VBN web-archive-org-7310 213 11 by by IN web-archive-org-7310 213 12 the the DT web-archive-org-7310 213 13 web web NN web-archive-org-7310 213 14 server server NN web-archive-org-7310 213 15 's 's POS web-archive-org-7310 213 16 robots.txt robots.txt NN web-archive-org-7310 213 17 file file NN web-archive-org-7310 213 18 , , , web-archive-org-7310 213 19 and and CC web-archive-org-7310 213 20 that that IN web-archive-org-7310 213 21 we -PRON- PRP web-archive-org-7310 213 22 therefore therefore RB web-archive-org-7310 213 23 refrained refrain VBD web-archive-org-7310 213 24 from from IN web-archive-org-7310 213 25 even even RB web-archive-org-7310 213 26 attempting attempt VBG web-archive-org-7310 213 27 to to TO web-archive-org-7310 213 28 download download VB web-archive-org-7310 213 29 . . . web-archive-org-7310 214 1 Again again RB web-archive-org-7310 214 2 , , , web-archive-org-7310 214 3 since since IN web-archive-org-7310 214 4 these these DT web-archive-org-7310 214 5 pages page NNS web-archive-org-7310 214 6 were be VBD web-archive-org-7310 214 7 not not RB web-archive-org-7310 214 8 excluded exclude VBN web-archive-org-7310 214 9 during during IN web-archive-org-7310 214 10 crawl crawl NNP web-archive-org-7310 214 11 1 1 CD web-archive-org-7310 214 12 , , , web-archive-org-7310 214 13 the the DT web-archive-org-7310 214 14 exclusion exclusion NN web-archive-org-7310 214 15 was be VBD web-archive-org-7310 214 16 imposed impose VBN web-archive-org-7310 214 17 later later RB web-archive-org-7310 214 18 . . . web-archive-org-7310 215 1 This this DT web-archive-org-7310 215 2 appears appear VBZ web-archive-org-7310 215 3 to to TO web-archive-org-7310 215 4 be be VB web-archive-org-7310 215 5 a a DT web-archive-org-7310 215 6 form form NN web-archive-org-7310 215 7 of of IN web-archive-org-7310 215 8 the the DT web-archive-org-7310 215 9 Heisenberg Heisenberg NNP web-archive-org-7310 215 10 effect effect NN web-archive-org-7310 215 11 , , , web-archive-org-7310 215 12 where where WRB web-archive-org-7310 215 13 the the DT web-archive-org-7310 215 14 existence existence NN web-archive-org-7310 215 15 of of IN web-archive-org-7310 215 16 an an DT web-archive-org-7310 215 17 observer observer NN web-archive-org-7310 215 18 ( ( -LRB- web-archive-org-7310 215 19 a a DT web-archive-org-7310 215 20 pesky pesky JJ web-archive-org-7310 215 21 web web NN web-archive-org-7310 215 22 crawler crawler NN web-archive-org-7310 215 23 that that WDT web-archive-org-7310 215 24 trots trot NNS web-archive-org-7310 215 25 by by IN web-archive-org-7310 215 26 every every DT web-archive-org-7310 215 27 week week NN web-archive-org-7310 215 28 ) ) -RRB- web-archive-org-7310 215 29 changes change VBZ web-archive-org-7310 215 30 the the DT web-archive-org-7310 215 31 behavior behavior NN web-archive-org-7310 215 32 of of IN web-archive-org-7310 215 33 the the DT web-archive-org-7310 215 34 observed observe VBN web-archive-org-7310 215 35 . . . web-archive-org-7310 216 1 As as IN web-archive-org-7310 216 2 one one PRP web-archive-org-7310 216 3 might may MD web-archive-org-7310 216 4 expect expect VB web-archive-org-7310 216 5 , , , web-archive-org-7310 216 6 Figure figure NN web-archive-org-7310 216 7   _SP web-archive-org-7310 216 8 4 4 CD web-archive-org-7310 216 9 bears bear VBZ web-archive-org-7310 216 10 out out RP web-archive-org-7310 216 11 that that IN web-archive-org-7310 216 12 the the DT web-archive-org-7310 216 13 lifetime lifetime NN web-archive-org-7310 216 14 of of IN web-archive-org-7310 216 15 a a DT web-archive-org-7310 216 16 URL url NN web-archive-org-7310 216 17 is be VBZ web-archive-org-7310 216 18 not not RB web-archive-org-7310 216 19 unlimited unlimited JJ web-archive-org-7310 216 20 . . . web-archive-org-7310 217 1 As as IN web-archive-org-7310 217 2 the the DT web-archive-org-7310 217 3 crawl crawl NN web-archive-org-7310 217 4 generations generation NNS web-archive-org-7310 217 5 increase increase VBP web-archive-org-7310 217 6 , , , web-archive-org-7310 217 7 more more JJR web-archive-org-7310 217 8 and and CC web-archive-org-7310 217 9 more more JJR web-archive-org-7310 217 10 URLs url NNS web-archive-org-7310 217 11 move move NN web-archive-org-7310 217 12 , , , web-archive-org-7310 217 13 become become VBP web-archive-org-7310 217 14 unreachable unreachable JJ web-archive-org-7310 217 15 , , , web-archive-org-7310 217 16 or or CC web-archive-org-7310 217 17 are be VBP web-archive-org-7310 217 18 blocked block VBN web-archive-org-7310 217 19 from from IN web-archive-org-7310 217 20 us -PRON- PRP web-archive-org-7310 217 21 . . . web-archive-org-7310 218 1 While while IN web-archive-org-7310 218 2 one one PRP web-archive-org-7310 218 3 would would MD web-archive-org-7310 218 4 expect expect VB web-archive-org-7310 218 5 a a DT web-archive-org-7310 218 6 geometric geometric JJ web-archive-org-7310 218 7 progression progression NN web-archive-org-7310 218 8 of of IN web-archive-org-7310 218 9 these these DT web-archive-org-7310 218 10 trends trend NNS web-archive-org-7310 218 11 , , , web-archive-org-7310 218 12 we -PRON- PRP web-archive-org-7310 218 13 did do VBD web-archive-org-7310 218 14 not not RB web-archive-org-7310 218 15 observe observe VB web-archive-org-7310 218 16 the the DT web-archive-org-7310 218 17 web web NN web-archive-org-7310 218 18 long long RB web-archive-org-7310 218 19 enough enough RB web-archive-org-7310 218 20 to to TO web-archive-org-7310 218 21 distinguish distinguish VB web-archive-org-7310 218 22 the the DT web-archive-org-7310 218 23 trend trend NN web-archive-org-7310 218 24 from from IN web-archive-org-7310 218 25 a a DT web-archive-org-7310 218 26 linear linear JJ web-archive-org-7310 218 27 progression progression NN web-archive-org-7310 218 28 . . . web-archive-org-7310 219 1 The the DT web-archive-org-7310 219 2 growth growth NN web-archive-org-7310 219 3 of of IN web-archive-org-7310 219 4 these these DT web-archive-org-7310 219 5 three three CD web-archive-org-7310 219 6 categories category NNS web-archive-org-7310 219 7 comes come VBZ web-archive-org-7310 219 8 at at IN web-archive-org-7310 219 9 the the DT web-archive-org-7310 219 10 expense expense NN web-archive-org-7310 219 11 of of IN web-archive-org-7310 219 12 the the DT web-archive-org-7310 219 13 200 200 CD web-archive-org-7310 219 14 category category NN web-archive-org-7310 219 15 ; ; , web-archive-org-7310 219 16 fluctuations fluctuation NNS web-archive-org-7310 219 17 in in IN web-archive-org-7310 219 18 the the DT web-archive-org-7310 219 19 share share NN web-archive-org-7310 219 20 of of IN web-archive-org-7310 219 21 the the DT web-archive-org-7310 219 22 network network NN web-archive-org-7310 219 23 category category NN web-archive-org-7310 219 24 appear appear VBP web-archive-org-7310 219 25 better well RBR web-archive-org-7310 219 26 correlated correlate VBN web-archive-org-7310 219 27 with with IN web-archive-org-7310 219 28 network network NN web-archive-org-7310 219 29 conditions condition NNS web-archive-org-7310 219 30 at at IN web-archive-org-7310 219 31 the the DT web-archive-org-7310 219 32 crawler crawler NNP web-archive-org-7310 219 33 side side NN web-archive-org-7310 219 34 . . . web-archive-org-7310 220 1 Figure figure NN web-archive-org-7310 220 2 4 4 CD web-archive-org-7310 220 3 : : : web-archive-org-7310 220 4 Distribution distribution NN web-archive-org-7310 220 5 of of IN web-archive-org-7310 220 6 HTTP HTTP VBN web-archive-org-7310 220 7 status status NN web-archive-org-7310 220 8 codes code NNS web-archive-org-7310 220 9 over over IN web-archive-org-7310 220 10 crawl crawl NN web-archive-org-7310 220 11 generations generation NNS web-archive-org-7310 220 12 . . . web-archive-org-7310 221 1 Figure figure NN web-archive-org-7310 221 2   _SP web-archive-org-7310 221 3 5 5 CD web-archive-org-7310 221 4 shows show VBZ web-archive-org-7310 221 5 only only RB web-archive-org-7310 221 6 the the DT web-archive-org-7310 221 7 successful successful JJ web-archive-org-7310 221 8 downloads download NNS web-archive-org-7310 221 9 as as IN web-archive-org-7310 221 10 a a DT web-archive-org-7310 221 11 percentage percentage NN web-archive-org-7310 221 12 of of IN web-archive-org-7310 221 13 all all DT web-archive-org-7310 221 14 download download NN web-archive-org-7310 221 15 attempts attempt NNS web-archive-org-7310 221 16 ( ( -LRB- web-archive-org-7310 221 17 including include VBG web-archive-org-7310 221 18 non non JJ web-archive-org-7310 221 19 - - NNS web-archive-org-7310 221 20 attempts attempt NNS web-archive-org-7310 221 21 due due IN web-archive-org-7310 221 22 to to IN web-archive-org-7310 221 23 robot robot NN web-archive-org-7310 221 24 exclusion exclusion NN web-archive-org-7310 221 25 rules rule NNS web-archive-org-7310 221 26 ) ) -RRB- web-archive-org-7310 221 27 , , , web-archive-org-7310 221 28 broken break VBN web-archive-org-7310 221 29 down down RP web-archive-org-7310 221 30 by by IN web-archive-org-7310 221 31 a a DT web-archive-org-7310 221 32 few few JJ web-archive-org-7310 221 33 selected select VBN web-archive-org-7310 221 34 top top JJ web-archive-org-7310 221 35 - - HYPH web-archive-org-7310 221 36 level level NN web-archive-org-7310 221 37 domains domain NNS web-archive-org-7310 221 38 . . . web-archive-org-7310 222 1 Note note VB web-archive-org-7310 222 2 that that IN web-archive-org-7310 222 3 the the DT web-archive-org-7310 222 4 y y NN web-archive-org-7310 222 5 - - HYPH web-archive-org-7310 222 6 axis axis NNP web-archive-org-7310 222 7 starts start VBZ web-archive-org-7310 222 8 at at IN web-archive-org-7310 222 9 80 80 CD web-archive-org-7310 222 10 % % NN web-archive-org-7310 222 11 , , , web-archive-org-7310 222 12 reflecting reflect VBG web-archive-org-7310 222 13 the the DT web-archive-org-7310 222 14 fact fact NN web-archive-org-7310 222 15 that that IN web-archive-org-7310 222 16 all all DT web-archive-org-7310 222 17 domains domain NNS web-archive-org-7310 222 18 in in IN web-archive-org-7310 222 19 all all DT web-archive-org-7310 222 20 generations generation NNS web-archive-org-7310 222 21 had have VBD web-archive-org-7310 222 22 at at IN web-archive-org-7310 222 23 least least JJS web-archive-org-7310 222 24 that that DT web-archive-org-7310 222 25 level level NN web-archive-org-7310 222 26 of of IN web-archive-org-7310 222 27 success success NN web-archive-org-7310 222 28 . . . web-archive-org-7310 223 1 In in IN web-archive-org-7310 223 2 general general JJ web-archive-org-7310 223 3 , , , web-archive-org-7310 223 4 pages page NNS web-archive-org-7310 223 5 in in IN web-archive-org-7310 223 6 .jp .jp NNP web-archive-org-7310 223 7 , , , web-archive-org-7310 223 8 .de .de . web-archive-org-7310 223 9 , , , web-archive-org-7310 223 10 and and CC web-archive-org-7310 223 11 .edu .edu NFP web-archive-org-7310 223 12 were be VBD web-archive-org-7310 223 13 consistently consistently RB web-archive-org-7310 223 14 more more RBR web-archive-org-7310 223 15 available available JJ web-archive-org-7310 223 16 than than IN web-archive-org-7310 223 17 pages page NNS web-archive-org-7310 223 18 in in IN web-archive-org-7310 223 19 .net .net . web-archive-org-7310 223 20 and and CC web-archive-org-7310 223 21 .com .com ADD web-archive-org-7310 223 22 . . . web-archive-org-7310 224 1 The the DT web-archive-org-7310 224 2 decline decline NN web-archive-org-7310 224 3 in in IN web-archive-org-7310 224 4 the the DT web-archive-org-7310 224 5 curves curve NNS web-archive-org-7310 224 6 bears bear VBZ web-archive-org-7310 224 7 out out RP web-archive-org-7310 224 8 the the DT web-archive-org-7310 224 9 limited limited JJ web-archive-org-7310 224 10 lifetime lifetime NN web-archive-org-7310 224 11 of of IN web-archive-org-7310 224 12 web web NN web-archive-org-7310 224 13 pages page NNS web-archive-org-7310 224 14 discussed discuss VBN web-archive-org-7310 224 15 above above RB web-archive-org-7310 224 16 . . . web-archive-org-7310 225 1 Figure figure NN web-archive-org-7310 225 2 5 5 CD web-archive-org-7310 225 3 : : : web-archive-org-7310 225 4 Distribution distribution NN web-archive-org-7310 225 5 of of IN web-archive-org-7310 225 6 successful successful JJ web-archive-org-7310 225 7 downloads download NNS web-archive-org-7310 225 8 over over IN web-archive-org-7310 225 9 crawl crawl NN web-archive-org-7310 225 10 generations generation NNS web-archive-org-7310 225 11 , , , web-archive-org-7310 225 12 broken break VBN web-archive-org-7310 225 13 down down RP web-archive-org-7310 225 14 by by IN web-archive-org-7310 225 15 selected select VBN web-archive-org-7310 225 16 top top JJ web-archive-org-7310 225 17 - - HYPH web-archive-org-7310 225 18 level level NN web-archive-org-7310 225 19 domains domain NNS web-archive-org-7310 225 20 . . . web-archive-org-7310 226 1 In in IN web-archive-org-7310 226 2 Figure figure NN web-archive-org-7310 226 3   _SP web-archive-org-7310 226 4 6 6 CD web-archive-org-7310 226 5 we -PRON- PRP web-archive-org-7310 226 6 tried try VBD web-archive-org-7310 226 7 another another DT web-archive-org-7310 226 8 approach approach NN web-archive-org-7310 226 9 for for IN web-archive-org-7310 226 10 viewing view VBG web-archive-org-7310 226 11 the the DT web-archive-org-7310 226 12 lifetime lifetime NN web-archive-org-7310 226 13 of of IN web-archive-org-7310 226 14 URLs URLs NNPS web-archive-org-7310 226 15 from from IN web-archive-org-7310 226 16 different different JJ web-archive-org-7310 226 17 domains domain NNS web-archive-org-7310 226 18 . . . web-archive-org-7310 227 1 Each each DT web-archive-org-7310 227 2 bar bar NN web-archive-org-7310 227 3 represents represent VBZ web-archive-org-7310 227 4 a a DT web-archive-org-7310 227 5 top top JJ web-archive-org-7310 227 6 - - HYPH web-archive-org-7310 227 7 level level NN web-archive-org-7310 227 8 domain domain NN web-archive-org-7310 227 9 ( ( -LRB- web-archive-org-7310 227 10 the the DT web-archive-org-7310 227 11 leftmost leftmost JJ web-archive-org-7310 227 12 bar bar NN web-archive-org-7310 227 13 represents represent VBZ web-archive-org-7310 227 14 the the DT web-archive-org-7310 227 15 entire entire JJ web-archive-org-7310 227 16 data data NN web-archive-org-7310 227 17 set set NN web-archive-org-7310 227 18 ) ) -RRB- web-archive-org-7310 227 19 . . . web-archive-org-7310 228 1 We -PRON- PRP web-archive-org-7310 228 2 grouped group VBD web-archive-org-7310 228 3 URLs url NNS web-archive-org-7310 228 4 by by IN web-archive-org-7310 228 5 the the DT web-archive-org-7310 228 6 crawl crawl JJ web-archive-org-7310 228 7 generation generation NN web-archive-org-7310 228 8 of of IN web-archive-org-7310 228 9 their -PRON- PRP$ web-archive-org-7310 228 10 last last JJ web-archive-org-7310 228 11 successful successful JJ web-archive-org-7310 228 12 retrieval retrieval NN web-archive-org-7310 228 13 , , , web-archive-org-7310 228 14 the the DT web-archive-org-7310 228 15 intuition intuition NN web-archive-org-7310 228 16 being be VBG web-archive-org-7310 228 17 that that IN web-archive-org-7310 228 18 a a DT web-archive-org-7310 228 19 URL url NN web-archive-org-7310 228 20 which which WDT web-archive-org-7310 228 21 could could MD web-archive-org-7310 228 22 not not RB web-archive-org-7310 228 23 be be VB web-archive-org-7310 228 24 downloaded download VBN web-archive-org-7310 228 25 after after IN web-archive-org-7310 228 26 some some DT web-archive-org-7310 228 27 point point NN web-archive-org-7310 228 28 is be VBZ web-archive-org-7310 228 29 likely likely JJ web-archive-org-7310 228 30 to to TO web-archive-org-7310 228 31 have have VB web-archive-org-7310 228 32 expired expire VBN web-archive-org-7310 228 33 . . . web-archive-org-7310 229 1 This this DT web-archive-org-7310 229 2 approach approach NN web-archive-org-7310 229 3 partitions partition VBZ web-archive-org-7310 229 4 URLs url NNS web-archive-org-7310 229 5 into into IN web-archive-org-7310 229 6 11 11 CD web-archive-org-7310 229 7 sets set NNS web-archive-org-7310 229 8 . . . web-archive-org-7310 230 1 Each each DT web-archive-org-7310 230 2 shaded shaded JJ web-archive-org-7310 230 3 region region NN web-archive-org-7310 230 4 of of IN web-archive-org-7310 230 5 each each DT web-archive-org-7310 230 6 bar bar NN web-archive-org-7310 230 7 represents represent VBZ web-archive-org-7310 230 8 the the DT web-archive-org-7310 230 9 relative relative JJ web-archive-org-7310 230 10 size size NN web-archive-org-7310 230 11 of of IN web-archive-org-7310 230 12 one one CD web-archive-org-7310 230 13 such such JJ web-archive-org-7310 230 14 set set NN web-archive-org-7310 230 15 . . . web-archive-org-7310 231 1 The the DT web-archive-org-7310 231 2 region region NN web-archive-org-7310 231 3 at at IN web-archive-org-7310 231 4 the the DT web-archive-org-7310 231 5 top top NN web-archive-org-7310 231 6 of of IN web-archive-org-7310 231 7 a a DT web-archive-org-7310 231 8 bar bar NN web-archive-org-7310 231 9 corresponds correspond NNS web-archive-org-7310 231 10 to to IN web-archive-org-7310 231 11 URLs url NNS web-archive-org-7310 231 12 that that WDT web-archive-org-7310 231 13 were be VBD web-archive-org-7310 231 14 consistently consistently RB web-archive-org-7310 231 15 unreachable unreachable JJ web-archive-org-7310 231 16 after after IN web-archive-org-7310 231 17 crawl crawl NNP web-archive-org-7310 231 18 1 1 CD web-archive-org-7310 231 19 ( ( -LRB- web-archive-org-7310 231 20 the the DT web-archive-org-7310 231 21 crawl crawl NN web-archive-org-7310 231 22 that that WDT web-archive-org-7310 231 23 defined define VBD web-archive-org-7310 231 24 the the DT web-archive-org-7310 231 25 set set NN web-archive-org-7310 231 26 of of IN web-archive-org-7310 231 27 URLs URLs NNPS web-archive-org-7310 231 28 ) ) -RRB- web-archive-org-7310 231 29 , , , web-archive-org-7310 231 30 while while IN web-archive-org-7310 231 31 the the DT web-archive-org-7310 231 32 region region NN web-archive-org-7310 231 33 at at IN web-archive-org-7310 231 34 the the DT web-archive-org-7310 231 35 bottom bottom NN web-archive-org-7310 231 36 of of IN web-archive-org-7310 231 37 a a DT web-archive-org-7310 231 38 bar bar NN web-archive-org-7310 231 39 corresponds correspond NNS web-archive-org-7310 231 40 to to IN web-archive-org-7310 231 41 URLs url NNS web-archive-org-7310 231 42 that that WDT web-archive-org-7310 231 43 were be VBD web-archive-org-7310 231 44 successfully successfully RB web-archive-org-7310 231 45 downloaded download VBN web-archive-org-7310 231 46 during during IN web-archive-org-7310 231 47 the the DT web-archive-org-7310 231 48 final final JJ web-archive-org-7310 231 49 crawl crawl NN web-archive-org-7310 231 50 . . . web-archive-org-7310 232 1 Note note VB web-archive-org-7310 232 2 that that IN web-archive-org-7310 232 3 the the DT web-archive-org-7310 232 4 y y NN web-archive-org-7310 232 5 - - HYPH web-archive-org-7310 232 6 axis axis NNP web-archive-org-7310 232 7 , , , web-archive-org-7310 232 8 which which WDT web-archive-org-7310 232 9 shows show VBZ web-archive-org-7310 232 10 the the DT web-archive-org-7310 232 11 percentage percentage NN web-archive-org-7310 232 12 breakdown breakdown NN web-archive-org-7310 232 13 of of IN web-archive-org-7310 232 14 the the DT web-archive-org-7310 232 15 total total NN web-archive-org-7310 232 16 , , , web-archive-org-7310 232 17 starts start VBZ web-archive-org-7310 232 18 at at IN web-archive-org-7310 232 19 75 75 CD web-archive-org-7310 232 20 % % NN web-archive-org-7310 232 21 , , , web-archive-org-7310 232 22 due due IN web-archive-org-7310 232 23 to to IN web-archive-org-7310 232 24 the the DT web-archive-org-7310 232 25 fact fact NN web-archive-org-7310 232 26 that that IN web-archive-org-7310 232 27 for for IN web-archive-org-7310 232 28 all all DT web-archive-org-7310 232 29 domains domain NNS web-archive-org-7310 232 30 considered consider VBN web-archive-org-7310 232 31 here here RB web-archive-org-7310 232 32 , , , web-archive-org-7310 232 33 more more JJR web-archive-org-7310 232 34 than than IN web-archive-org-7310 232 35 75 75 CD web-archive-org-7310 232 36 % % NN web-archive-org-7310 232 37 of of IN web-archive-org-7310 232 38 all all DT web-archive-org-7310 232 39 URLs url NNS web-archive-org-7310 232 40 were be VBD web-archive-org-7310 232 41 still still RB web-archive-org-7310 232 42 reachable reachable JJ web-archive-org-7310 232 43 during during IN web-archive-org-7310 232 44 the the DT web-archive-org-7310 232 45 final final JJ web-archive-org-7310 232 46 crawl crawl NN web-archive-org-7310 232 47 . . . web-archive-org-7310 233 1 Looking look VBG web-archive-org-7310 233 2 at at IN web-archive-org-7310 233 3 the the DT web-archive-org-7310 233 4 " " `` web-archive-org-7310 233 5 all all DT web-archive-org-7310 233 6 domains domain NNS web-archive-org-7310 233 7 " " '' web-archive-org-7310 233 8 bar bar NN web-archive-org-7310 233 9 , , , web-archive-org-7310 233 10 it -PRON- PRP web-archive-org-7310 233 11 can can MD web-archive-org-7310 233 12 be be VB web-archive-org-7310 233 13 seen see VBN web-archive-org-7310 233 14 that that IN web-archive-org-7310 233 15 88 88 CD web-archive-org-7310 233 16 % % NN web-archive-org-7310 233 17 of of IN web-archive-org-7310 233 18 all all DT web-archive-org-7310 233 19 URLs url NNS web-archive-org-7310 233 20 were be VBD web-archive-org-7310 233 21 still still RB web-archive-org-7310 233 22 available available JJ web-archive-org-7310 233 23 during during IN web-archive-org-7310 233 24 the the DT web-archive-org-7310 233 25 final final JJ web-archive-org-7310 233 26 crawl crawl NN web-archive-org-7310 233 27 . . . web-archive-org-7310 234 1 For for IN web-archive-org-7310 234 2 most most JJS web-archive-org-7310 234 3 domains domain NNS web-archive-org-7310 234 4 , , , web-archive-org-7310 234 5 the the DT web-archive-org-7310 234 6 " " `` web-archive-org-7310 234 7 OK ok UH web-archive-org-7310 234 8 at at IN web-archive-org-7310 234 9 crawl crawl NN web-archive-org-7310 234 10 10 10 CD web-archive-org-7310 234 11 " " `` web-archive-org-7310 234 12 regions region NNS web-archive-org-7310 234 13 are be VBP web-archive-org-7310 234 14 larger large JJR web-archive-org-7310 234 15 than than IN web-archive-org-7310 234 16 regions region NNS web-archive-org-7310 234 17 for for IN web-archive-org-7310 234 18 preceding precede VBG web-archive-org-7310 234 19 crawls crawl NNS web-archive-org-7310 234 20 . . . web-archive-org-7310 235 1 This this DT web-archive-org-7310 235 2 makes make VBZ web-archive-org-7310 235 3 intuitive intuitive JJ web-archive-org-7310 235 4 sense sense NN web-archive-org-7310 235 5 : : : web-archive-org-7310 235 6 it -PRON- PRP web-archive-org-7310 235 7 represents represent VBZ web-archive-org-7310 235 8 documents document NNS web-archive-org-7310 235 9 that that WDT web-archive-org-7310 235 10 could could MD web-archive-org-7310 235 11 not not RB web-archive-org-7310 235 12 be be VB web-archive-org-7310 235 13 retrieved retrieve VBN web-archive-org-7310 235 14 during during IN web-archive-org-7310 235 15 the the DT web-archive-org-7310 235 16 final final JJ web-archive-org-7310 235 17 crawl crawl NN web-archive-org-7310 235 18 , , , web-archive-org-7310 235 19 but but CC web-archive-org-7310 235 20 might may MD web-archive-org-7310 235 21 well well RB web-archive-org-7310 235 22 come come VB web-archive-org-7310 235 23 back back RB web-archive-org-7310 235 24 to to IN web-archive-org-7310 235 25 life life NN web-archive-org-7310 235 26 in in IN web-archive-org-7310 235 27 the the DT web-archive-org-7310 235 28 future future NN web-archive-org-7310 235 29 . . . web-archive-org-7310 236 1 Other other JJ web-archive-org-7310 236 2 than than IN web-archive-org-7310 236 3 that that DT web-archive-org-7310 236 4 , , , web-archive-org-7310 236 5 we -PRON- PRP web-archive-org-7310 236 6 see see VBP web-archive-org-7310 236 7 no no DT web-archive-org-7310 236 8 discernible discernible JJ web-archive-org-7310 236 9 patterns pattern NNS web-archive-org-7310 236 10 between between IN web-archive-org-7310 236 11 the the DT web-archive-org-7310 236 12 lengths length NNS web-archive-org-7310 236 13 of of IN web-archive-org-7310 236 14 regions region NNS web-archive-org-7310 236 15 within within IN web-archive-org-7310 236 16 a a DT web-archive-org-7310 236 17 bar bar NN web-archive-org-7310 236 18 . . . web-archive-org-7310 237 1 Looking look VBG web-archive-org-7310 237 2 across across IN web-archive-org-7310 237 3 domains domain NNS web-archive-org-7310 237 4 , , , web-archive-org-7310 237 5 we -PRON- PRP web-archive-org-7310 237 6 observe observe VBP web-archive-org-7310 237 7 that that IN web-archive-org-7310 237 8 web web NN web-archive-org-7310 237 9 pages page NNS web-archive-org-7310 237 10 in in IN web-archive-org-7310 237 11 China China NNP web-archive-org-7310 237 12 expire expire VBP web-archive-org-7310 237 13 sooner soon RBR web-archive-org-7310 237 14 than than IN web-archive-org-7310 237 15 average average JJ web-archive-org-7310 237 16 , , , web-archive-org-7310 237 17 as as IN web-archive-org-7310 237 18 do do VBP web-archive-org-7310 237 19 pages page NNS web-archive-org-7310 237 20 in in IN web-archive-org-7310 237 21 .com .com . web-archive-org-7310 237 22 and and CC web-archive-org-7310 237 23 .net .net . web-archive-org-7310 237 24 . . . web-archive-org-7310 238 1 Figure figure NN web-archive-org-7310 238 2 6 6 CD web-archive-org-7310 238 3 : : : web-archive-org-7310 238 4 Breakdown breakdown NN web-archive-org-7310 238 5 showing show VBG web-archive-org-7310 238 6 in in IN web-archive-org-7310 238 7 which which WDT web-archive-org-7310 238 8 crawl crawl NN web-archive-org-7310 238 9 a a DT web-archive-org-7310 238 10 web web NN web-archive-org-7310 238 11 page page NN web-archive-org-7310 238 12 was be VBD web-archive-org-7310 238 13 last last JJ web-archive-org-7310 238 14 successfully successfully RB web-archive-org-7310 238 15 downloaded download VBN web-archive-org-7310 238 16 , , , web-archive-org-7310 238 17 broken break VBN web-archive-org-7310 238 18 down down RP web-archive-org-7310 238 19 by by IN web-archive-org-7310 238 20 TLD tld NN web-archive-org-7310 238 21 . . . web-archive-org-7310 239 1 The the DT web-archive-org-7310 239 2 remaining remain VBG web-archive-org-7310 239 3 figures figure NNS web-archive-org-7310 239 4 display display VBP web-archive-org-7310 239 5 information information NN web-archive-org-7310 239 6 about about IN web-archive-org-7310 239 7 the the DT web-archive-org-7310 239 8 amount amount NN web-archive-org-7310 239 9 of of IN web-archive-org-7310 239 10 change change NN web-archive-org-7310 239 11 in in IN web-archive-org-7310 239 12 a a DT web-archive-org-7310 239 13 document document NN web-archive-org-7310 239 14 between between IN web-archive-org-7310 239 15 two two CD web-archive-org-7310 239 16 successive successive JJ web-archive-org-7310 239 17 successful successful JJ web-archive-org-7310 239 18 downloads download NNS web-archive-org-7310 239 19 . . . web-archive-org-7310 240 1 Figure figure NN web-archive-org-7310 240 2   _SP web-archive-org-7310 240 3 7 7 CD web-archive-org-7310 240 4 shows show VBZ web-archive-org-7310 240 5 a a DT web-archive-org-7310 240 6 fine fine RB web-archive-org-7310 240 7 - - HYPH web-archive-org-7310 240 8 grained grain VBN web-archive-org-7310 240 9 illustration illustration NN web-archive-org-7310 240 10 of of IN web-archive-org-7310 240 11 change change NN web-archive-org-7310 240 12 amount amount NN web-archive-org-7310 240 13 , , , web-archive-org-7310 240 14 independent independent JJ web-archive-org-7310 240 15 of of IN web-archive-org-7310 240 16 other other JJ web-archive-org-7310 240 17 factors factor NNS web-archive-org-7310 240 18 . . . web-archive-org-7310 241 1 We -PRON- PRP web-archive-org-7310 241 2 partition partition VBP web-archive-org-7310 241 3 the the DT web-archive-org-7310 241 4 set set NN web-archive-org-7310 241 5 of of IN web-archive-org-7310 241 6 all all DT web-archive-org-7310 241 7 pairs pair NNS web-archive-org-7310 241 8 of of IN web-archive-org-7310 241 9 successive successive JJ web-archive-org-7310 241 10 successfully successfully RB web-archive-org-7310 241 11 retrieved retrieve VBN web-archive-org-7310 241 12 pages page NNS web-archive-org-7310 241 13 into into IN web-archive-org-7310 241 14 85 85 CD web-archive-org-7310 241 15 subsets subset NNS web-archive-org-7310 241 16 , , , web-archive-org-7310 241 17 based base VBN web-archive-org-7310 241 18 on on IN web-archive-org-7310 241 19 how how WRB web-archive-org-7310 241 20 many many JJ web-archive-org-7310 241 21 pre pre NNS web-archive-org-7310 241 22 - - NNS web-archive-org-7310 241 23 images image VBZ web-archive-org-7310 241 24 the the DT web-archive-org-7310 241 25 two two CD web-archive-org-7310 241 26 documents document NNS web-archive-org-7310 241 27 in in IN web-archive-org-7310 241 28 each each DT web-archive-org-7310 241 29 pair pair NN web-archive-org-7310 241 30 share share NN web-archive-org-7310 241 31 . . . web-archive-org-7310 242 1 Subset subset NN web-archive-org-7310 242 2 0 0 NFP web-archive-org-7310 242 3 contains contain VBZ web-archive-org-7310 242 4 the the DT web-archive-org-7310 242 5 pairs pair NNS web-archive-org-7310 242 6 with with IN web-archive-org-7310 242 7 no no DT web-archive-org-7310 242 8 common common JJ web-archive-org-7310 242 9 pre pre NNS web-archive-org-7310 242 10 - - NNS web-archive-org-7310 242 11 images image NNS web-archive-org-7310 242 12 , , , web-archive-org-7310 242 13 subset subset VBP web-archive-org-7310 242 14 84 84 CD web-archive-org-7310 242 15 the the DT web-archive-org-7310 242 16 ones one NNS web-archive-org-7310 242 17 with with IN web-archive-org-7310 242 18 all all DT web-archive-org-7310 242 19 common common JJ web-archive-org-7310 242 20 pre pre NNS web-archive-org-7310 242 21 - - NNS web-archive-org-7310 242 22 images image NNS web-archive-org-7310 242 23 , , , web-archive-org-7310 242 24 and and CC web-archive-org-7310 242 25 subset subset VBP web-archive-org-7310 242 26 85 85 CD web-archive-org-7310 242 27 the the DT web-archive-org-7310 242 28 ones one NNS web-archive-org-7310 242 29 that that WDT web-archive-org-7310 242 30 also also RB web-archive-org-7310 242 31 agree agree VBP web-archive-org-7310 242 32 in in IN web-archive-org-7310 242 33 their -PRON- PRP$ web-archive-org-7310 242 34 document document NN web-archive-org-7310 242 35 checksums checksum NNS web-archive-org-7310 242 36 ( ( -LRB- web-archive-org-7310 242 37 strongly strongly RB web-archive-org-7310 242 38 suggesting suggest VBG web-archive-org-7310 242 39 that that IN web-archive-org-7310 242 40 the the DT web-archive-org-7310 242 41 documents document NNS web-archive-org-7310 242 42 are be VBP web-archive-org-7310 242 43 identical identical JJ web-archive-org-7310 242 44 as as IN web-archive-org-7310 242 45 byte byte NN web-archive-org-7310 242 46 sequences sequence NNS web-archive-org-7310 242 47 ) ) -RRB- web-archive-org-7310 242 48 . . . web-archive-org-7310 243 1 The the DT web-archive-org-7310 243 2 x x NN web-archive-org-7310 243 3 - - : web-archive-org-7310 243 4 axis axis RB web-archive-org-7310 243 5 shows show VBZ web-archive-org-7310 243 6 the the DT web-archive-org-7310 243 7 85 85 CD web-archive-org-7310 243 8 subsets subset NNS web-archive-org-7310 243 9 , , , web-archive-org-7310 243 10 and and CC web-archive-org-7310 243 11 the the DT web-archive-org-7310 243 12 y y NN web-archive-org-7310 243 13 - - HYPH web-archive-org-7310 243 14 axis axis NN web-archive-org-7310 243 15 shows show VBZ web-archive-org-7310 243 16 percentages percentage NNS web-archive-org-7310 243 17 . . . web-archive-org-7310 244 1 The the DT web-archive-org-7310 244 2 lower low JJR web-archive-org-7310 244 3 curve curve NN web-archive-org-7310 244 4 ( ( -LRB- web-archive-org-7310 244 5 visible visible JJ web-archive-org-7310 244 6 only only RB web-archive-org-7310 244 7 at at IN web-archive-org-7310 244 8 the the DT web-archive-org-7310 244 9 extreme extreme JJ web-archive-org-7310 244 10 left left NN web-archive-org-7310 244 11 and and CC web-archive-org-7310 244 12 right right NN web-archive-org-7310 244 13 ) ) -RRB- web-archive-org-7310 244 14 depicts depict VBZ web-archive-org-7310 244 15 what what WDT web-archive-org-7310 244 16 percentage percentage NN web-archive-org-7310 244 17 of of IN web-archive-org-7310 244 18 all all DT web-archive-org-7310 244 19 documents document NNS web-archive-org-7310 244 20 fall fall VBP web-archive-org-7310 244 21 into into IN web-archive-org-7310 244 22 a a DT web-archive-org-7310 244 23 given give VBN web-archive-org-7310 244 24 subset subset NN web-archive-org-7310 244 25 . . . web-archive-org-7310 245 1 The the DT web-archive-org-7310 245 2 curve curve NN web-archive-org-7310 245 3 above above RB web-archive-org-7310 245 4 shows show VBZ web-archive-org-7310 245 5 the the DT web-archive-org-7310 245 6 cumulative cumulative JJ web-archive-org-7310 245 7 percentage percentage NN web-archive-org-7310 245 8 distribution distribution NN web-archive-org-7310 245 9 . . . web-archive-org-7310 246 1 Each each DT web-archive-org-7310 246 2 point point NN web-archive-org-7310 246 3 ( ( -LRB- web-archive-org-7310 246 4 x x NN web-archive-org-7310 246 5 , , , web-archive-org-7310 246 6 y y NNP web-archive-org-7310 246 7 ) ) -RRB- web-archive-org-7310 246 8 on on IN web-archive-org-7310 246 9 the the DT web-archive-org-7310 246 10 curve curve NN web-archive-org-7310 246 11 indicates indicate VBZ web-archive-org-7310 246 12 that that IN web-archive-org-7310 246 13 y y NNP web-archive-org-7310 246 14 percent percent NN web-archive-org-7310 246 15 of of IN web-archive-org-7310 246 16 all all DT web-archive-org-7310 246 17 pairs pair NNS web-archive-org-7310 246 18 had have VBD web-archive-org-7310 246 19 at at IN web-archive-org-7310 246 20 least least JJS web-archive-org-7310 246 21 x x DT web-archive-org-7310 246 22 pre pre VBN web-archive-org-7310 246 23 - - NNS web-archive-org-7310 246 24 images image NNS web-archive-org-7310 246 25 in in IN web-archive-org-7310 246 26 common common JJ web-archive-org-7310 246 27 . . . web-archive-org-7310 247 1 As as IN web-archive-org-7310 247 2 can can MD web-archive-org-7310 247 3 be be VB web-archive-org-7310 247 4 seen see VBN web-archive-org-7310 247 5 ( ( -LRB- web-archive-org-7310 247 6 if if IN web-archive-org-7310 247 7 your -PRON- PRP$ web-archive-org-7310 247 8 eyes eye NNS web-archive-org-7310 247 9 are be VBP web-archive-org-7310 247 10 sharp sharp JJ web-archive-org-7310 247 11 enough enough RB web-archive-org-7310 247 12 ) ) -RRB- web-archive-org-7310 247 13 , , , web-archive-org-7310 247 14 65.2 65.2 CD web-archive-org-7310 247 15 % % NN web-archive-org-7310 247 16 of of IN web-archive-org-7310 247 17 all all DT web-archive-org-7310 247 18 page page NN web-archive-org-7310 247 19 pairs pair NNS web-archive-org-7310 247 20 do do VBP web-archive-org-7310 247 21 n't not RB web-archive-org-7310 247 22 differ differ VB web-archive-org-7310 247 23 at at RB web-archive-org-7310 247 24 all all RB web-archive-org-7310 247 25 . . . web-archive-org-7310 248 1 Another another DT web-archive-org-7310 248 2 9.2 9.2 CD web-archive-org-7310 248 3 % % NN web-archive-org-7310 248 4 differ differ VBP web-archive-org-7310 248 5 only only RB web-archive-org-7310 248 6 in in IN web-archive-org-7310 248 7 their -PRON- PRP$ web-archive-org-7310 248 8 checksum checksum NN web-archive-org-7310 248 9 , , , web-archive-org-7310 248 10 but but CC web-archive-org-7310 248 11 have have VBP web-archive-org-7310 248 12 all all DT web-archive-org-7310 248 13 common common JJ web-archive-org-7310 248 14 pre pre NNS web-archive-org-7310 248 15 - - NNS web-archive-org-7310 248 16 images image NNS web-archive-org-7310 248 17 ( ( -LRB- web-archive-org-7310 248 18 suggesting suggest VBG web-archive-org-7310 248 19 that that IN web-archive-org-7310 248 20 only only RB web-archive-org-7310 248 21 the the DT web-archive-org-7310 248 22 HTML html NN web-archive-org-7310 248 23 markup markup NN web-archive-org-7310 248 24 , , , web-archive-org-7310 248 25 which which WDT web-archive-org-7310 248 26 was be VBD web-archive-org-7310 248 27 removed remove VBN web-archive-org-7310 248 28 before before IN web-archive-org-7310 248 29 computing compute VBG web-archive-org-7310 248 30 pre pre NN web-archive-org-7310 248 31 - - NNS web-archive-org-7310 248 32 images image NNS web-archive-org-7310 248 33 , , , web-archive-org-7310 248 34 has have VBZ web-archive-org-7310 248 35 changed change VBN web-archive-org-7310 248 36 ) ) -RRB- web-archive-org-7310 248 37 . . . web-archive-org-7310 249 1 Figure figure NN web-archive-org-7310 249 2 7 7 CD web-archive-org-7310 249 3 : : : web-archive-org-7310 249 4 Distribution distribution NN web-archive-org-7310 249 5 of of IN web-archive-org-7310 249 6 change change NN web-archive-org-7310 249 7 . . . web-archive-org-7310 250 1 Figure figure NN web-archive-org-7310 250 2   _SP web-archive-org-7310 250 3 8 8 CD web-archive-org-7310 250 4 magnifies magnifie NNS web-archive-org-7310 250 5 the the DT web-archive-org-7310 250 6 lower low JJR web-archive-org-7310 250 7 curve curve NN web-archive-org-7310 250 8 from from IN web-archive-org-7310 250 9 Figure figure NN web-archive-org-7310 250 10   _SP web-archive-org-7310 250 11 7 7 CD web-archive-org-7310 250 12 , , , web-archive-org-7310 250 13 making make VBG web-archive-org-7310 250 14 it -PRON- PRP web-archive-org-7310 250 15 easy easy JJ web-archive-org-7310 250 16 to to TO web-archive-org-7310 250 17 see see VB web-archive-org-7310 250 18 that that IN web-archive-org-7310 250 19 all all DT web-archive-org-7310 250 20 other other JJ web-archive-org-7310 250 21 change change NN web-archive-org-7310 250 22 buckets bucket NNS web-archive-org-7310 250 23 contain contain VBP web-archive-org-7310 250 24 less less JJR web-archive-org-7310 250 25 than than IN web-archive-org-7310 250 26 2 2 CD web-archive-org-7310 250 27 % % NN web-archive-org-7310 250 28 of of IN web-archive-org-7310 250 29 all all DT web-archive-org-7310 250 30 page page NN web-archive-org-7310 250 31 pairs pair NNS web-archive-org-7310 250 32 , , , web-archive-org-7310 250 33 and and CC web-archive-org-7310 250 34 that that IN web-archive-org-7310 250 35 all all DT web-archive-org-7310 250 36 buckets bucket NNS web-archive-org-7310 250 37 with with IN web-archive-org-7310 250 38 fewer few JJR web-archive-org-7310 250 39 than than IN web-archive-org-7310 250 40 79 79 CD web-archive-org-7310 250 41 common common JJ web-archive-org-7310 250 42 pre pre NNS web-archive-org-7310 250 43 - - NNS web-archive-org-7310 250 44 images image NNS web-archive-org-7310 250 45 ( ( -LRB- web-archive-org-7310 250 46 representing represent VBG web-archive-org-7310 250 47 document document NN web-archive-org-7310 250 48 pairs pair NNS web-archive-org-7310 250 49 that that WDT web-archive-org-7310 250 50 are be VBP web-archive-org-7310 250 51 less less JJR web-archive-org-7310 250 52 than than IN web-archive-org-7310 250 53 94 94 CD web-archive-org-7310 250 54 % % NN web-archive-org-7310 250 55 similar similar JJ web-archive-org-7310 250 56 ) ) -RRB- web-archive-org-7310 250 57 contain contain VBP web-archive-org-7310 250 58 less less JJR web-archive-org-7310 250 59 than than IN web-archive-org-7310 250 60 1 1 CD web-archive-org-7310 250 61 % % NN web-archive-org-7310 250 62 . . . web-archive-org-7310 251 1 For for IN web-archive-org-7310 251 2 the the DT web-archive-org-7310 251 3 most most JJS web-archive-org-7310 251 4 part part NN web-archive-org-7310 251 5 , , , web-archive-org-7310 251 6 the the DT web-archive-org-7310 251 7 curve curve NN web-archive-org-7310 251 8 is be VBZ web-archive-org-7310 251 9 monotonically monotonically RB web-archive-org-7310 251 10 decreasing decrease VBG web-archive-org-7310 251 11 , , , web-archive-org-7310 251 12 but but CC web-archive-org-7310 251 13 this this DT web-archive-org-7310 251 14 pattern pattern NN web-archive-org-7310 251 15 is be VBZ web-archive-org-7310 251 16 broken break VBN web-archive-org-7310 251 17 as as IN web-archive-org-7310 251 18 we -PRON- PRP web-archive-org-7310 251 19 get get VBP web-archive-org-7310 251 20 to to IN web-archive-org-7310 251 21 the the DT web-archive-org-7310 251 22 buckets bucket NNS web-archive-org-7310 251 23 containing contain VBG web-archive-org-7310 251 24 documents document NNS web-archive-org-7310 251 25 that that WDT web-archive-org-7310 251 26 underwent undergo VBD web-archive-org-7310 251 27 extreme extreme JJ web-archive-org-7310 251 28 change change NN web-archive-org-7310 251 29 . . . web-archive-org-7310 252 1 Bucket Bucket NNP web-archive-org-7310 252 2 0 0 CD web-archive-org-7310 252 3 , , , web-archive-org-7310 252 4 which which WDT web-archive-org-7310 252 5 represents represent VBZ web-archive-org-7310 252 6 complete complete JJ web-archive-org-7310 252 7 dissimilarity dissimilarity NN web-archive-org-7310 252 8 , , , web-archive-org-7310 252 9 accounts account VBZ web-archive-org-7310 252 10 for for IN web-archive-org-7310 252 11 0.8 0.8 CD web-archive-org-7310 252 12 % % NN web-archive-org-7310 252 13 , , , web-archive-org-7310 252 14 over over IN web-archive-org-7310 252 15 ten ten CD web-archive-org-7310 252 16 times time NNS web-archive-org-7310 252 17 the the DT web-archive-org-7310 252 18 level level NN web-archive-org-7310 252 19 of of IN web-archive-org-7310 252 20 bucket bucket NN web-archive-org-7310 252 21 7 7 CD web-archive-org-7310 252 22 , , , web-archive-org-7310 252 23 the the DT web-archive-org-7310 252 24 smallest small JJS web-archive-org-7310 252 25 bucket bucket NN web-archive-org-7310 252 26 . . . web-archive-org-7310 253 1 We -PRON- PRP web-archive-org-7310 253 2 will will MD web-archive-org-7310 253 3 revisit revisit VB web-archive-org-7310 253 4 this this DT web-archive-org-7310 253 5 phenomenon phenomenon NN web-archive-org-7310 253 6 later later RB web-archive-org-7310 253 7 on on RB web-archive-org-7310 253 8 . . . web-archive-org-7310 254 1 Figure figure NN web-archive-org-7310 254 2 8 8 CD web-archive-org-7310 254 3 : : : web-archive-org-7310 254 4 Distribution distribution NN web-archive-org-7310 254 5 of of IN web-archive-org-7310 254 6 change change NN web-archive-org-7310 254 7 , , , web-archive-org-7310 254 8 scaled scale VBN web-archive-org-7310 254 9 to to TO web-archive-org-7310 254 10 show show VB web-archive-org-7310 254 11 low low JJ web-archive-org-7310 254 12 - - HYPH web-archive-org-7310 254 13 percentage percentage NN web-archive-org-7310 254 14 categories category NNS web-archive-org-7310 254 15 . . . web-archive-org-7310 255 1 Figure figure NN web-archive-org-7310 255 2 9 9 CD web-archive-org-7310 255 3 : : : web-archive-org-7310 255 4 Distribution distribution NN web-archive-org-7310 255 5 of of IN web-archive-org-7310 255 6 change change NN web-archive-org-7310 255 7 , , , web-archive-org-7310 255 8 scaled scale VBN web-archive-org-7310 255 9 to to TO web-archive-org-7310 255 10 show show VB web-archive-org-7310 255 11 low low JJ web-archive-org-7310 255 12 - - HYPH web-archive-org-7310 255 13 percentage percentage NN web-archive-org-7310 255 14 categories category NNS web-archive-org-7310 255 15 , , , web-archive-org-7310 255 16 after after IN web-archive-org-7310 255 17 excluding exclude VBG web-archive-org-7310 255 18 automatically automatically RB web-archive-org-7310 255 19 generated generate VBN web-archive-org-7310 255 20 keyword keyword NNP web-archive-org-7310 255 21 - - HYPH web-archive-org-7310 255 22 spam spam NNP web-archive-org-7310 255 23 documents document NNS web-archive-org-7310 255 24 . . . web-archive-org-7310 256 1 We -PRON- PRP web-archive-org-7310 256 2 next next RB web-archive-org-7310 256 3 tried try VBD web-archive-org-7310 256 4 to to TO web-archive-org-7310 256 5 determine determine VB web-archive-org-7310 256 6 if if IN web-archive-org-7310 256 7 there there EX web-archive-org-7310 256 8 are be VBP web-archive-org-7310 256 9 any any DT web-archive-org-7310 256 10 attributes attribute NNS web-archive-org-7310 256 11 of of IN web-archive-org-7310 256 12 a a DT web-archive-org-7310 256 13 document document NN web-archive-org-7310 256 14 that that WDT web-archive-org-7310 256 15 help help VBP web-archive-org-7310 256 16 to to TO web-archive-org-7310 256 17 predict predict VB web-archive-org-7310 256 18 its -PRON- PRP$ web-archive-org-7310 256 19 rate rate NN web-archive-org-7310 256 20 and and CC web-archive-org-7310 256 21 degree degree NN web-archive-org-7310 256 22 of of IN web-archive-org-7310 256 23 change change NN web-archive-org-7310 256 24 . . . web-archive-org-7310 257 1 In in IN web-archive-org-7310 257 2 Figure figure NN web-archive-org-7310 257 3   _SP web-archive-org-7310 257 4 10 10 CD web-archive-org-7310 257 5 , , , web-archive-org-7310 257 6 we -PRON- PRP web-archive-org-7310 257 7 examine examine VBP web-archive-org-7310 257 8 the the DT web-archive-org-7310 257 9 relation relation NN web-archive-org-7310 257 10 between between IN web-archive-org-7310 257 11 top top JJ web-archive-org-7310 257 12 - - HYPH web-archive-org-7310 257 13 level level NN web-archive-org-7310 257 14 domain domain NN web-archive-org-7310 257 15 and and CC web-archive-org-7310 257 16 change change NN web-archive-org-7310 257 17 . . . web-archive-org-7310 258 1 The the DT web-archive-org-7310 258 2 figure figure NN web-archive-org-7310 258 3 shows show VBZ web-archive-org-7310 258 4 a a DT web-archive-org-7310 258 5 bar bar NN web-archive-org-7310 258 6 for for IN web-archive-org-7310 258 7 all all DT web-archive-org-7310 258 8 page page NN web-archive-org-7310 258 9 pairs pair NNS web-archive-org-7310 258 10 and and CC web-archive-org-7310 258 11 several several JJ web-archive-org-7310 258 12 more more JJR web-archive-org-7310 258 13 for for IN web-archive-org-7310 258 14 selected select VBN web-archive-org-7310 258 15 domains domain NNS web-archive-org-7310 258 16 . . . web-archive-org-7310 259 1 Each each DT web-archive-org-7310 259 2 bar bar NN web-archive-org-7310 259 3 is be VBZ web-archive-org-7310 259 4 divided divide VBN web-archive-org-7310 259 5 into into IN web-archive-org-7310 259 6 6 6 CD web-archive-org-7310 259 7 regions region NNS web-archive-org-7310 259 8 , , , web-archive-org-7310 259 9 corresponding correspond VBG web-archive-org-7310 259 10 to to IN web-archive-org-7310 259 11 the the DT web-archive-org-7310 259 12 following following JJ web-archive-org-7310 259 13 six six CD web-archive-org-7310 259 14 change change NN web-archive-org-7310 259 15 clusters cluster NNS web-archive-org-7310 259 16 : : : web-archive-org-7310 259 17 complete complete JJ web-archive-org-7310 259 18 change change NN web-archive-org-7310 259 19 ( ( -LRB- web-archive-org-7310 259 20 0 0 CD web-archive-org-7310 259 21 common common JJ web-archive-org-7310 259 22 pre pre NNS web-archive-org-7310 259 23 - - NNS web-archive-org-7310 259 24 images image NNS web-archive-org-7310 259 25 ) ) -RRB- web-archive-org-7310 259 26 , , , web-archive-org-7310 259 27 large large JJ web-archive-org-7310 259 28 change change NN web-archive-org-7310 259 29 ( ( -LRB- web-archive-org-7310 259 30 1 1 CD web-archive-org-7310 259 31 - - SYM web-archive-org-7310 259 32 28 28 CD web-archive-org-7310 259 33 common common JJ web-archive-org-7310 259 34 pre pre NNS web-archive-org-7310 259 35 - - NNS web-archive-org-7310 259 36 images image NNS web-archive-org-7310 259 37 ) ) -RRB- web-archive-org-7310 259 38 , , , web-archive-org-7310 259 39 medium medium JJ web-archive-org-7310 259 40 change change NN web-archive-org-7310 259 41 ( ( -LRB- web-archive-org-7310 259 42 29 29 CD web-archive-org-7310 259 43 - - SYM web-archive-org-7310 259 44 56 56 CD web-archive-org-7310 259 45 common common JJ web-archive-org-7310 259 46 pre pre NNS web-archive-org-7310 259 47 - - NNS web-archive-org-7310 259 48 images image NNS web-archive-org-7310 259 49 ) ) -RRB- web-archive-org-7310 259 50 , , , web-archive-org-7310 259 51 small small JJ web-archive-org-7310 259 52 change change NN web-archive-org-7310 259 53 ( ( -LRB- web-archive-org-7310 259 54 57 57 CD web-archive-org-7310 259 55 - - SYM web-archive-org-7310 259 56 83 83 CD web-archive-org-7310 259 57 common common JJ web-archive-org-7310 259 58 pre pre NNS web-archive-org-7310 259 59 - - NNS web-archive-org-7310 259 60 images image NNS web-archive-org-7310 259 61 ) ) -RRB- web-archive-org-7310 259 62 , , , web-archive-org-7310 259 63 no no DT web-archive-org-7310 259 64 text text NN web-archive-org-7310 259 65 change change NN web-archive-org-7310 259 66 ( ( -LRB- web-archive-org-7310 259 67 84 84 CD web-archive-org-7310 259 68 common common JJ web-archive-org-7310 259 69 pre pre NNS web-archive-org-7310 259 70 - - NNS web-archive-org-7310 259 71 images image NNS web-archive-org-7310 259 72 ) ) -RRB- web-archive-org-7310 259 73 , , , web-archive-org-7310 259 74 and and CC web-archive-org-7310 259 75 no no DT web-archive-org-7310 259 76 change change NN web-archive-org-7310 259 77 ( ( -LRB- web-archive-org-7310 259 78 subset subset VBN web-archive-org-7310 259 79 85 85 CD web-archive-org-7310 259 80 : : : web-archive-org-7310 259 81 84 84 CD web-archive-org-7310 259 82 common common JJ web-archive-org-7310 259 83 pre pre NNS web-archive-org-7310 259 84 - - NNS web-archive-org-7310 259 85 images image NNS web-archive-org-7310 259 86 and and CC web-archive-org-7310 259 87 a a DT web-archive-org-7310 259 88 common common JJ web-archive-org-7310 259 89 checksum checksum NN web-archive-org-7310 259 90 ) ) -RRB- web-archive-org-7310 259 91 . . . web-archive-org-7310 260 1 The the DT web-archive-org-7310 260 2 top top JJ web-archive-org-7310 260 3 region region NN web-archive-org-7310 260 4 depicts depict VBZ web-archive-org-7310 260 5 the the DT web-archive-org-7310 260 6 complete complete JJ web-archive-org-7310 260 7 change change NN web-archive-org-7310 260 8 cluster cluster NN web-archive-org-7310 260 9 , , , web-archive-org-7310 260 10 the the DT web-archive-org-7310 260 11 one one NN web-archive-org-7310 260 12 at at IN web-archive-org-7310 260 13 the the DT web-archive-org-7310 260 14 bottom bottom NN web-archive-org-7310 260 15 the the DT web-archive-org-7310 260 16 no no DT web-archive-org-7310 260 17 change change NN web-archive-org-7310 260 18 cluster cluster NN web-archive-org-7310 260 19 . . . web-archive-org-7310 261 1 The the DT web-archive-org-7310 261 2 y y NN web-archive-org-7310 261 3 - - HYPH web-archive-org-7310 261 4 axis axis NNP web-archive-org-7310 261 5 shows show VBZ web-archive-org-7310 261 6 the the DT web-archive-org-7310 261 7 percentage percentage NN web-archive-org-7310 261 8 . . . web-archive-org-7310 262 1 Figure figure NN web-archive-org-7310 262 2 10 10 CD web-archive-org-7310 262 3 : : : web-archive-org-7310 262 4 Clustered clustered JJ web-archive-org-7310 262 5 rates rate NNS web-archive-org-7310 262 6 of of IN web-archive-org-7310 262 7 change change NN web-archive-org-7310 262 8 , , , web-archive-org-7310 262 9 broken break VBN web-archive-org-7310 262 10 down down RP web-archive-org-7310 262 11 by by IN web-archive-org-7310 262 12 selected select VBN web-archive-org-7310 262 13 top top JJ web-archive-org-7310 262 14 - - HYPH web-archive-org-7310 262 15 level level NN web-archive-org-7310 262 16 domains domain NNS web-archive-org-7310 262 17 . . . web-archive-org-7310 263 1 We -PRON- PRP web-archive-org-7310 263 2 observe observe VBP web-archive-org-7310 263 3 significant significant JJ web-archive-org-7310 263 4 differences difference NNS web-archive-org-7310 263 5 between between IN web-archive-org-7310 263 6 top top JJ web-archive-org-7310 263 7 - - HYPH web-archive-org-7310 263 8 level level NN web-archive-org-7310 263 9 domains domain NNS web-archive-org-7310 263 10 , , , web-archive-org-7310 263 11 confirming confirm VBG web-archive-org-7310 263 12 earlier early JJR web-archive-org-7310 263 13 observations observation NNS web-archive-org-7310 263 14 by by IN web-archive-org-7310 263 15 Cho Cho NNP web-archive-org-7310 263 16 and and CC web-archive-org-7310 263 17 Garcia Garcia NNP web-archive-org-7310 263 18 - - HYPH web-archive-org-7310 263 19 Molina Molina NNP web-archive-org-7310 263 20   _SP web-archive-org-7310 263 21 [ [ -LRB- web-archive-org-7310 263 22 5 5 CD web-archive-org-7310 263 23 ] ] -RRB- web-archive-org-7310 263 24 . . . web-archive-org-7310 264 1 In in IN web-archive-org-7310 264 2 the the DT web-archive-org-7310 264 3 .com .com CD web-archive-org-7310 264 4 domain domain NN web-archive-org-7310 264 5 , , , web-archive-org-7310 264 6 pages page NNS web-archive-org-7310 264 7 change change VBP web-archive-org-7310 264 8 more more RBR web-archive-org-7310 264 9 frequently frequently RB web-archive-org-7310 264 10 than than IN web-archive-org-7310 264 11 in in IN web-archive-org-7310 264 12 the the DT web-archive-org-7310 264 13 .gov .gov . web-archive-org-7310 264 14 and and CC web-archive-org-7310 264 15 .edu .edu CD web-archive-org-7310 264 16 domains domain NNS web-archive-org-7310 264 17 . . . web-archive-org-7310 265 1 We -PRON- PRP web-archive-org-7310 265 2 were be VBD web-archive-org-7310 265 3 surprised surprised JJ web-archive-org-7310 265 4 to to TO web-archive-org-7310 265 5 see see VB web-archive-org-7310 265 6 that that IN web-archive-org-7310 265 7 pages page NNS web-archive-org-7310 265 8 in in IN web-archive-org-7310 265 9 .de .de NNP web-archive-org-7310 265 10 , , , web-archive-org-7310 265 11 the the DT web-archive-org-7310 265 12 German german JJ web-archive-org-7310 265 13 domain domain NN web-archive-org-7310 265 14 , , , web-archive-org-7310 265 15 exhibit exhibit VB web-archive-org-7310 265 16 a a DT web-archive-org-7310 265 17 significantly significantly RB web-archive-org-7310 265 18 higher high JJR web-archive-org-7310 265 19 rate rate NN web-archive-org-7310 265 20 and and CC web-archive-org-7310 265 21 degree degree NN web-archive-org-7310 265 22 of of IN web-archive-org-7310 265 23 change change NN web-archive-org-7310 265 24 than than IN web-archive-org-7310 265 25 those those DT web-archive-org-7310 265 26 in in IN web-archive-org-7310 265 27 any any DT web-archive-org-7310 265 28 other other JJ web-archive-org-7310 265 29 domain domain NN web-archive-org-7310 265 30 . . . web-archive-org-7310 266 1 27 27 CD web-archive-org-7310 266 2 % % NN web-archive-org-7310 266 3 of of IN web-archive-org-7310 266 4 the the DT web-archive-org-7310 266 5 pages page NNS web-archive-org-7310 266 6 we -PRON- PRP web-archive-org-7310 266 7 sampled sample VBD web-archive-org-7310 266 8 from from IN web-archive-org-7310 266 9 .de .de NFP web-archive-org-7310 266 10 underwent undergo VBD web-archive-org-7310 266 11 a a DT web-archive-org-7310 266 12 large large JJ web-archive-org-7310 266 13 or or CC web-archive-org-7310 266 14 complete complete JJ web-archive-org-7310 266 15 change change NN web-archive-org-7310 266 16 every every DT web-archive-org-7310 266 17 week week NN web-archive-org-7310 266 18 , , , web-archive-org-7310 266 19 compared compare VBN web-archive-org-7310 266 20 with with IN web-archive-org-7310 266 21 3 3 CD web-archive-org-7310 266 22 % % NN web-archive-org-7310 266 23 for for IN web-archive-org-7310 266 24 the the DT web-archive-org-7310 266 25 web web NN web-archive-org-7310 266 26 as as IN web-archive-org-7310 266 27 a a DT web-archive-org-7310 266 28 whole whole NN web-archive-org-7310 266 29 . . . web-archive-org-7310 267 1 Even even RB web-archive-org-7310 267 2 taking take VBG web-archive-org-7310 267 3 the the DT web-archive-org-7310 267 4 fabled fabled JJ web-archive-org-7310 267 5 German german JJ web-archive-org-7310 267 6 industriousness industriousness NN web-archive-org-7310 267 7 into into IN web-archive-org-7310 267 8 account account NN web-archive-org-7310 267 9 , , , web-archive-org-7310 267 10 these these DT web-archive-org-7310 267 11 numbers number NNS web-archive-org-7310 267 12 were be VBD web-archive-org-7310 267 13 hard hard JJ web-archive-org-7310 267 14 to to TO web-archive-org-7310 267 15 explain explain VB web-archive-org-7310 267 16 . . . web-archive-org-7310 268 1 In in IN web-archive-org-7310 268 2 order order NN web-archive-org-7310 268 3 to to TO web-archive-org-7310 268 4 shed shed VB web-archive-org-7310 268 5 light light NN web-archive-org-7310 268 6 on on IN web-archive-org-7310 268 7 the the DT web-archive-org-7310 268 8 issue issue NN web-archive-org-7310 268 9 , , , web-archive-org-7310 268 10 we -PRON- PRP web-archive-org-7310 268 11 turned turn VBD web-archive-org-7310 268 12 to to IN web-archive-org-7310 268 13 our -PRON- PRP$ web-archive-org-7310 268 14 sampled sample VBN web-archive-org-7310 268 15 documents document NNS web-archive-org-7310 268 16 , , , web-archive-org-7310 268 17 selecting select VBG web-archive-org-7310 268 18 documents document NNS web-archive-org-7310 268 19 from from IN web-archive-org-7310 268 20 Germany Germany NNP web-archive-org-7310 268 21 with with IN web-archive-org-7310 268 22 high high JJ web-archive-org-7310 268 23 change change NN web-archive-org-7310 268 24 rate rate NN web-archive-org-7310 268 25 . . . web-archive-org-7310 269 1 Careful careful JJ web-archive-org-7310 269 2 examination examination NN web-archive-org-7310 269 3 of of IN web-archive-org-7310 269 4 the the DT web-archive-org-7310 269 5 first first JJ web-archive-org-7310 269 6 few few JJ web-archive-org-7310 269 7 pages page NNS web-archive-org-7310 269 8 revealed reveal VBD web-archive-org-7310 269 9 more more RBR web-archive-org-7310 269 10 than than IN web-archive-org-7310 269 11 we -PRON- PRP web-archive-org-7310 269 12 cared care VBD web-archive-org-7310 269 13 to to TO web-archive-org-7310 269 14 see see VB web-archive-org-7310 269 15 : : : web-archive-org-7310 269 16 of of IN web-archive-org-7310 269 17 the the DT web-archive-org-7310 269 18 first first JJ web-archive-org-7310 269 19 half half JJ web-archive-org-7310 269 20 dozen dozen NN web-archive-org-7310 269 21 pages page NNS web-archive-org-7310 269 22 we -PRON- PRP web-archive-org-7310 269 23 examined examine VBD web-archive-org-7310 269 24 , , , web-archive-org-7310 269 25 all all DT web-archive-org-7310 269 26 but but IN web-archive-org-7310 269 27 one one CD web-archive-org-7310 269 28 contained contain VBN web-archive-org-7310 269 29 disjoint disjoint NN web-archive-org-7310 269 30 , , , web-archive-org-7310 269 31 but but CC web-archive-org-7310 269 32 perfectly perfectly RB web-archive-org-7310 269 33 grammatical grammatical JJ web-archive-org-7310 269 34 phrases phrase NNS web-archive-org-7310 269 35 of of IN web-archive-org-7310 269 36 an an DT web-archive-org-7310 269 37 adult adult NN web-archive-org-7310 269 38 nature nature NN web-archive-org-7310 269 39 together together RB web-archive-org-7310 269 40 with with IN web-archive-org-7310 269 41 a a DT web-archive-org-7310 269 42 redirection redirection NN web-archive-org-7310 269 43 to to IN web-archive-org-7310 269 44 an an DT web-archive-org-7310 269 45 adult adult NN web-archive-org-7310 269 46 web web NN web-archive-org-7310 269 47 site site NN web-archive-org-7310 269 48 . . . web-archive-org-7310 270 1 It -PRON- PRP web-archive-org-7310 270 2 soon soon RB web-archive-org-7310 270 3 became become VBD web-archive-org-7310 270 4 clear clear JJ web-archive-org-7310 270 5 that that IN web-archive-org-7310 270 6 the the DT web-archive-org-7310 270 7 phrases phrase NNS web-archive-org-7310 270 8 were be VBD web-archive-org-7310 270 9 automatically automatically RB web-archive-org-7310 270 10 generated generate VBN web-archive-org-7310 270 11 on on IN web-archive-org-7310 270 12 the the DT web-archive-org-7310 270 13 fly fly NN web-archive-org-7310 270 14 , , , web-archive-org-7310 270 15 for for IN web-archive-org-7310 270 16 the the DT web-archive-org-7310 270 17 purpose purpose NN web-archive-org-7310 270 18 of of IN web-archive-org-7310 270 19 " " `` web-archive-org-7310 270 20 stuffing stuff VBG web-archive-org-7310 270 21 " " '' web-archive-org-7310 270 22 search search NN web-archive-org-7310 270 23 engines engine NNS web-archive-org-7310 270 24 such such JJ web-archive-org-7310 270 25 as as IN web-archive-org-7310 270 26 Google Google NNP web-archive-org-7310 270 27 with with IN web-archive-org-7310 270 28 topical topical JJ web-archive-org-7310 270 29 keywords keyword NNS web-archive-org-7310 270 30 surrounded surround VBN web-archive-org-7310 270 31 by by IN web-archive-org-7310 270 32 sensible sensible NN web-archive-org-7310 270 33 - - HYPH web-archive-org-7310 270 34 looking look VBG web-archive-org-7310 270 35 context context NN web-archive-org-7310 270 36 , , , web-archive-org-7310 270 37 in in IN web-archive-org-7310 270 38 order order NN web-archive-org-7310 270 39 to to TO web-archive-org-7310 270 40 draw draw VB web-archive-org-7310 270 41 visitors visitor NNS web-archive-org-7310 270 42 to to IN web-archive-org-7310 270 43 the the DT web-archive-org-7310 270 44 adult adult NN web-archive-org-7310 270 45 web web NN web-archive-org-7310 270 46 site site NN web-archive-org-7310 270 47 . . . web-archive-org-7310 271 1 Upon upon IN web-archive-org-7310 271 2 further further JJ web-archive-org-7310 271 3 investigation investigation NN web-archive-org-7310 271 4 , , , web-archive-org-7310 271 5 we -PRON- PRP web-archive-org-7310 271 6 discovered discover VBD web-archive-org-7310 271 7 that that IN web-archive-org-7310 271 8 our -PRON- PRP$ web-archive-org-7310 271 9 data data NN web-archive-org-7310 271 10 set set NN web-archive-org-7310 271 11 contained contain VBD web-archive-org-7310 271 12 1.03 1.03 CD web-archive-org-7310 271 13 million million CD web-archive-org-7310 271 14 URLs url NNS web-archive-org-7310 271 15 drawn draw VBN web-archive-org-7310 271 16 from from IN web-archive-org-7310 271 17 116,654 116,654 CD web-archive-org-7310 271 18 hosts host NNS web-archive-org-7310 271 19 ( ( -LRB- web-archive-org-7310 271 20 4,745 4,745 CD web-archive-org-7310 271 21 of of IN web-archive-org-7310 271 22 them -PRON- PRP web-archive-org-7310 271 23 being be VBG web-archive-org-7310 271 24 outside outside IN web-archive-org-7310 271 25 the the DT web-archive-org-7310 271 26 .de .de JJ web-archive-org-7310 271 27 domain domain NN web-archive-org-7310 271 28 ) ) -RRB- web-archive-org-7310 271 29 , , , web-archive-org-7310 271 30 which which WDT web-archive-org-7310 271 31 all all DT web-archive-org-7310 271 32 resolved resolve VBD web-archive-org-7310 271 33 to to IN web-archive-org-7310 271 34 a a DT web-archive-org-7310 271 35 single single JJ web-archive-org-7310 271 36 IP IP NNP web-archive-org-7310 271 37 address address NN web-archive-org-7310 271 38 . . . web-archive-org-7310 272 1 This this DT web-archive-org-7310 272 2 machine machine NN web-archive-org-7310 272 3 is be VBZ web-archive-org-7310 272 4 serving serve VBG web-archive-org-7310 272 5 up up RP web-archive-org-7310 272 6 over over IN web-archive-org-7310 272 7 15 15 CD web-archive-org-7310 272 8 % % NN web-archive-org-7310 272 9 of of IN web-archive-org-7310 272 10 the the DT web-archive-org-7310 272 11 .de .de JJ web-archive-org-7310 272 12 URLs url NNS web-archive-org-7310 272 13 in in IN web-archive-org-7310 272 14 our -PRON- PRP$ web-archive-org-7310 272 15 data data NN web-archive-org-7310 272 16 set set NN web-archive-org-7310 272 17 ! ! . web-archive-org-7310 273 1 We -PRON- PRP web-archive-org-7310 273 2 speculate speculate VBP web-archive-org-7310 273 3 that that IN web-archive-org-7310 273 4 the the DT web-archive-org-7310 273 5 purpose purpose NN web-archive-org-7310 273 6 of of IN web-archive-org-7310 273 7 using use VBG web-archive-org-7310 273 8 that that IN web-archive-org-7310 273 9 many many JJ web-archive-org-7310 273 10 distinct distinct JJ web-archive-org-7310 273 11 host host NN web-archive-org-7310 273 12 names name NNS web-archive-org-7310 273 13 as as IN web-archive-org-7310 273 14 a a DT web-archive-org-7310 273 15 front front NN web-archive-org-7310 273 16 to to IN web-archive-org-7310 273 17 a a DT web-archive-org-7310 273 18 single single JJ web-archive-org-7310 273 19 server server NN web-archive-org-7310 273 20 is be VBZ web-archive-org-7310 273 21 to to TO web-archive-org-7310 273 22 circumvent circumvent VB web-archive-org-7310 273 23 the the DT web-archive-org-7310 273 24 politeness politeness NN web-archive-org-7310 273 25 policies policy NNS web-archive-org-7310 273 26 that that WDT web-archive-org-7310 273 27 limit limit VBP web-archive-org-7310 273 28 the the DT web-archive-org-7310 273 29 number number NN web-archive-org-7310 273 30 of of IN web-archive-org-7310 273 31 pages page NNS web-archive-org-7310 273 32 a a DT web-archive-org-7310 273 33 web web NN web-archive-org-7310 273 34 crawler crawler NN web-archive-org-7310 273 35 will will MD web-archive-org-7310 273 36 attempt attempt VB web-archive-org-7310 273 37 to to TO web-archive-org-7310 273 38 download download VB web-archive-org-7310 273 39 from from IN web-archive-org-7310 273 40 any any DT web-archive-org-7310 273 41 given give VBN web-archive-org-7310 273 42 host host NN web-archive-org-7310 273 43 in in IN web-archive-org-7310 273 44 a a DT web-archive-org-7310 273 45 given give VBN web-archive-org-7310 273 46 time time NN web-archive-org-7310 273 47 interval interval NN web-archive-org-7310 273 48 , , , web-archive-org-7310 273 49 and and CC web-archive-org-7310 273 50 also also RB web-archive-org-7310 273 51 to to TO web-archive-org-7310 273 52 trick trick VB web-archive-org-7310 273 53 link link NN web-archive-org-7310 273 54 - - HYPH web-archive-org-7310 273 55 based base VBN web-archive-org-7310 273 56 ranking rank VBG web-archive-org-7310 273 57 algorithms algorithm NNS web-archive-org-7310 273 58 such such JJ web-archive-org-7310 273 59 as as IN web-archive-org-7310 273 60 PageRank PageRank NNP web-archive-org-7310 273 61 into into IN web-archive-org-7310 273 62 believing believe VBG web-archive-org-7310 273 63 that that IN web-archive-org-7310 273 64 links link NNS web-archive-org-7310 273 65 to to IN web-archive-org-7310 273 66 other other JJ web-archive-org-7310 273 67 pages page NNS web-archive-org-7310 273 68 on on IN web-archive-org-7310 273 69 apparently apparently RB web-archive-org-7310 273 70 different different JJ web-archive-org-7310 273 71 hosts host NNS web-archive-org-7310 273 72 are be VBP web-archive-org-7310 273 73 non non JJ web-archive-org-7310 273 74 - - JJ web-archive-org-7310 273 75 nepotistic nepotistic JJ web-archive-org-7310 273 76 , , , web-archive-org-7310 273 77 thereby thereby RB web-archive-org-7310 273 78 inflating inflate VBG web-archive-org-7310 273 79 the the DT web-archive-org-7310 273 80 ranking ranking NN web-archive-org-7310 273 81 of of IN web-archive-org-7310 273 82 the the DT web-archive-org-7310 273 83 pages page NNS web-archive-org-7310 273 84 in in IN web-archive-org-7310 273 85 the the DT web-archive-org-7310 273 86 clique clique NN web-archive-org-7310 273 87 . . . web-archive-org-7310 274 1 After after IN web-archive-org-7310 274 2 this this DT web-archive-org-7310 274 3 discovery discovery NN web-archive-org-7310 274 4 , , , web-archive-org-7310 274 5 we -PRON- PRP web-archive-org-7310 274 6 set set VBD web-archive-org-7310 274 7 out out RP web-archive-org-7310 274 8 to to TO web-archive-org-7310 274 9 explore explore VB web-archive-org-7310 274 10 if if IN web-archive-org-7310 274 11 there there EX web-archive-org-7310 274 12 were be VBD web-archive-org-7310 274 13 other other JJ web-archive-org-7310 274 14 such such JJ web-archive-org-7310 274 15 servers server NNS web-archive-org-7310 274 16 in in IN web-archive-org-7310 274 17 our -PRON- PRP$ web-archive-org-7310 274 18 data data NN web-archive-org-7310 274 19 set set NN web-archive-org-7310 274 20 . . . web-archive-org-7310 275 1 We -PRON- PRP web-archive-org-7310 275 2 resolved resolve VBD web-archive-org-7310 275 3 the the DT web-archive-org-7310 275 4 symbolic symbolic JJ web-archive-org-7310 275 5 host host NN web-archive-org-7310 275 6 names name NNS web-archive-org-7310 275 7 of of IN web-archive-org-7310 275 8 all all PDT web-archive-org-7310 275 9 the the DT web-archive-org-7310 275 10 URLs url NNS web-archive-org-7310 275 11 in in IN web-archive-org-7310 275 12 our -PRON- PRP$ web-archive-org-7310 275 13 data data NN web-archive-org-7310 275 14 set set NN web-archive-org-7310 275 15 , , , web-archive-org-7310 275 16 and and CC web-archive-org-7310 275 17 singled single VBD web-archive-org-7310 275 18 out out RP web-archive-org-7310 275 19 each each DT web-archive-org-7310 275 20 IP IP NNP web-archive-org-7310 275 21 address address NN web-archive-org-7310 275 22 with with IN web-archive-org-7310 275 23 more more JJR web-archive-org-7310 275 24 than than IN web-archive-org-7310 275 25 a a DT web-archive-org-7310 275 26 thousand thousand CD web-archive-org-7310 275 27 symbolic symbolic JJ web-archive-org-7310 275 28 host host NN web-archive-org-7310 275 29 names name NNS web-archive-org-7310 275 30 mapping map VBG web-archive-org-7310 275 31 to to IN web-archive-org-7310 275 32 it -PRON- PRP web-archive-org-7310 275 33 . . . web-archive-org-7310 276 1 There there EX web-archive-org-7310 276 2 were be VBD web-archive-org-7310 276 3 213 213 CD web-archive-org-7310 276 4 such such JJ web-archive-org-7310 276 5 IP IP NNP web-archive-org-7310 276 6 addresses address NNS web-archive-org-7310 276 7 , , , web-archive-org-7310 276 8 78 78 CD web-archive-org-7310 276 9 of of IN web-archive-org-7310 276 10 which which WDT web-archive-org-7310 276 11 proved prove VBD web-archive-org-7310 276 12 to to TO web-archive-org-7310 276 13 be be VB web-archive-org-7310 276 14 of of IN web-archive-org-7310 276 15 a a DT web-archive-org-7310 276 16 similar similar JJ web-archive-org-7310 276 17 nature nature NN web-archive-org-7310 276 18 as as IN web-archive-org-7310 276 19 the the DT web-archive-org-7310 276 20 site site NN web-archive-org-7310 276 21 that that WDT web-archive-org-7310 276 22 triggered trigger VBD web-archive-org-7310 276 23 our -PRON- PRP$ web-archive-org-7310 276 24 investigation investigation NN web-archive-org-7310 276 25 . . . web-archive-org-7310 277 1 We -PRON- PRP web-archive-org-7310 277 2 excluded exclude VBD web-archive-org-7310 277 3 all all DT web-archive-org-7310 277 4 URLs url NNS web-archive-org-7310 277 5 on on IN web-archive-org-7310 277 6 the the DT web-archive-org-7310 277 7 443,038 443,038 CD web-archive-org-7310 277 8 hosts host NNS web-archive-org-7310 277 9 that that WDT web-archive-org-7310 277 10 resolved resolve VBD web-archive-org-7310 277 11 to to IN web-archive-org-7310 277 12 one one CD web-archive-org-7310 277 13 of of IN web-archive-org-7310 277 14 the the DT web-archive-org-7310 277 15 78 78 CD web-archive-org-7310 277 16 identified identify VBN web-archive-org-7310 277 17 IP IP NNP web-archive-org-7310 277 18 addresses address NNS web-archive-org-7310 277 19 , , , web-archive-org-7310 277 20 and and CC web-archive-org-7310 277 21 reran reran VB web-archive-org-7310 277 22 the the DT web-archive-org-7310 277 23 analysis analysis NN web-archive-org-7310 277 24 that that WDT web-archive-org-7310 277 25 produced produce VBD web-archive-org-7310 277 26 Figure figure NN web-archive-org-7310 277 27   _SP web-archive-org-7310 277 28 10 10 CD web-archive-org-7310 277 29 . . . web-archive-org-7310 278 1 This this DT web-archive-org-7310 278 2 eliminated eliminate VBD web-archive-org-7310 278 3 about about RB web-archive-org-7310 278 4 60 60 CD web-archive-org-7310 278 5 % % NN web-archive-org-7310 278 6 of of IN web-archive-org-7310 278 7 the the DT web-archive-org-7310 278 8 excessive excessive JJ web-archive-org-7310 278 9 large large JJ web-archive-org-7310 278 10 and and CC web-archive-org-7310 278 11 complete complete JJ web-archive-org-7310 278 12 change change NN web-archive-org-7310 278 13 in in IN web-archive-org-7310 278 14 .de .de NN web-archive-org-7310 278 15 . . . web-archive-org-7310 279 1 The the DT web-archive-org-7310 279 2 adjusted adjusted JJ web-archive-org-7310 279 3 distribution distribution NN web-archive-org-7310 279 4 is be VBZ web-archive-org-7310 279 5 shown show VBN web-archive-org-7310 279 6 in in IN web-archive-org-7310 279 7 Figure figure NN web-archive-org-7310 279 8   _SP web-archive-org-7310 279 9 11 11 CD web-archive-org-7310 279 10 . . . web-archive-org-7310 280 1 Continued continue VBN web-archive-org-7310 280 2 investigation investigation NN web-archive-org-7310 280 3 of of IN web-archive-org-7310 280 4 the the DT web-archive-org-7310 280 5 excessive excessive JJ web-archive-org-7310 280 6 change change NN web-archive-org-7310 280 7 found find VBD web-archive-org-7310 280 8 that that IN web-archive-org-7310 280 9 automatically automatically RB web-archive-org-7310 280 10 generated generate VBN web-archive-org-7310 280 11 pornographic pornographic JJ web-archive-org-7310 280 12 content content NN web-archive-org-7310 280 13 accounts account NNS web-archive-org-7310 280 14 for for IN web-archive-org-7310 280 15 much much JJ web-archive-org-7310 280 16 of of IN web-archive-org-7310 280 17 the the DT web-archive-org-7310 280 18 remainder remainder NN web-archive-org-7310 280 19 . . . web-archive-org-7310 281 1 Figure figure NN web-archive-org-7310 281 2 11 11 CD web-archive-org-7310 281 3 : : : web-archive-org-7310 281 4 Clustered clustered JJ web-archive-org-7310 281 5 rates rate NNS web-archive-org-7310 281 6 of of IN web-archive-org-7310 281 7 change change NN web-archive-org-7310 281 8 , , , web-archive-org-7310 281 9 broken break VBN web-archive-org-7310 281 10 down down RP web-archive-org-7310 281 11 by by IN web-archive-org-7310 281 12 selected select VBN web-archive-org-7310 281 13 top top JJ web-archive-org-7310 281 14 - - HYPH web-archive-org-7310 281 15 level level NN web-archive-org-7310 281 16 domains domain NNS web-archive-org-7310 281 17 , , , web-archive-org-7310 281 18 after after IN web-archive-org-7310 281 19 excluding exclude VBG web-archive-org-7310 281 20 automatically automatically RB web-archive-org-7310 281 21 generated generate VBN web-archive-org-7310 281 22 keyword keyword NNP web-archive-org-7310 281 23 - - HYPH web-archive-org-7310 281 24 spam spam NNP web-archive-org-7310 281 25 documents document NNS web-archive-org-7310 281 26 . . . web-archive-org-7310 282 1 In in IN web-archive-org-7310 282 2 Figure figure NN web-archive-org-7310 282 3   _SP web-archive-org-7310 282 4 12 12 CD web-archive-org-7310 282 5 , , , web-archive-org-7310 282 6 we -PRON- PRP web-archive-org-7310 282 7 look look VBP web-archive-org-7310 282 8 at at IN web-archive-org-7310 282 9 the the DT web-archive-org-7310 282 10 same same JJ web-archive-org-7310 282 11 data datum NNS web-archive-org-7310 282 12 , , , web-archive-org-7310 282 13 but but CC web-archive-org-7310 282 14 omit omit VB web-archive-org-7310 282 15 all all DT web-archive-org-7310 282 16 pairs pair NNS web-archive-org-7310 282 17 of of IN web-archive-org-7310 282 18 documents document NNS web-archive-org-7310 282 19 with with IN web-archive-org-7310 282 20 no no DT web-archive-org-7310 282 21 change change NN web-archive-org-7310 282 22 . . . web-archive-org-7310 283 1 Other other JJ web-archive-org-7310 283 2 than than IN web-archive-org-7310 283 3 Germany Germany NNP web-archive-org-7310 283 4 ( ( -LRB- web-archive-org-7310 283 5 and and CC web-archive-org-7310 283 6 to to IN web-archive-org-7310 283 7 a a DT web-archive-org-7310 283 8 lesser less JJR web-archive-org-7310 283 9 extent extent NN web-archive-org-7310 283 10 China China NNP web-archive-org-7310 283 11 and and CC web-archive-org-7310 283 12 Japan Japan NNP web-archive-org-7310 283 13 ) ) -RRB- web-archive-org-7310 283 14 , , , web-archive-org-7310 283 15 there there EX web-archive-org-7310 283 16 is be VBZ web-archive-org-7310 283 17 remarkably remarkably RB web-archive-org-7310 283 18 little little JJ web-archive-org-7310 283 19 difference difference NN web-archive-org-7310 283 20 between between IN web-archive-org-7310 283 21 the the DT web-archive-org-7310 283 22 various various JJ web-archive-org-7310 283 23 top top JJ web-archive-org-7310 283 24 - - HYPH web-archive-org-7310 283 25 level level NN web-archive-org-7310 283 26 domains domain NNS web-archive-org-7310 283 27 . . . web-archive-org-7310 284 1 Our -PRON- PRP$ web-archive-org-7310 284 2 conclusions conclusion NNS web-archive-org-7310 284 3 are be VBP web-archive-org-7310 284 4 twofold twofold JJ web-archive-org-7310 284 5 : : : web-archive-org-7310 284 6 First first RB web-archive-org-7310 284 7 , , , web-archive-org-7310 284 8 pornography pornography NN web-archive-org-7310 284 9 continues continue VBZ web-archive-org-7310 284 10 to to TO web-archive-org-7310 284 11 skew skew VB web-archive-org-7310 284 12 our -PRON- PRP$ web-archive-org-7310 284 13 results result NNS web-archive-org-7310 284 14 . . . web-archive-org-7310 285 1 Second second JJ web-archive-org-7310 285 2 , , , web-archive-org-7310 285 3 our -PRON- PRP$ web-archive-org-7310 285 4 shingling shingling NN web-archive-org-7310 285 5 technique technique NN web-archive-org-7310 285 6 is be VBZ web-archive-org-7310 285 7 not not RB web-archive-org-7310 285 8 well well RB web-archive-org-7310 285 9 adapted adapted JJ web-archive-org-7310 285 10 to to IN web-archive-org-7310 285 11 writing write VBG web-archive-org-7310 285 12 systems system NNS web-archive-org-7310 285 13 like like IN web-archive-org-7310 285 14 Chinese Chinese NNPS web-archive-org-7310 285 15 or or CC web-archive-org-7310 285 16 Kanji Kanji NNPS web-archive-org-7310 285 17 that that WDT web-archive-org-7310 285 18 do do VBP web-archive-org-7310 285 19 not not RB web-archive-org-7310 285 20 employ employ VB web-archive-org-7310 285 21 inter inter JJ web-archive-org-7310 285 22 - - JJ web-archive-org-7310 285 23 word word JJ web-archive-org-7310 285 24 spacing spacing NN web-archive-org-7310 285 25 , , , web-archive-org-7310 285 26 which which WDT web-archive-org-7310 285 27 in in IN web-archive-org-7310 285 28 turn turn NN web-archive-org-7310 285 29 causes cause VBZ web-archive-org-7310 285 30 documents document NNS web-archive-org-7310 285 31 to to TO web-archive-org-7310 285 32 have have VB web-archive-org-7310 285 33 a a DT web-archive-org-7310 285 34 very very RB web-archive-org-7310 285 35 small small JJ web-archive-org-7310 285 36 number number NN web-archive-org-7310 285 37 of of IN web-archive-org-7310 285 38 shingles shingle NNS web-archive-org-7310 285 39 , , , web-archive-org-7310 285 40 which which WDT web-archive-org-7310 285 41 means mean VBZ web-archive-org-7310 285 42 that that IN web-archive-org-7310 285 43 any any DT web-archive-org-7310 285 44 change change NN web-archive-org-7310 285 45 is be VBZ web-archive-org-7310 285 46 considered consider VBN web-archive-org-7310 285 47 significant significant JJ web-archive-org-7310 285 48 . . . web-archive-org-7310 286 1 Figure figure NN web-archive-org-7310 286 2 12 12 CD web-archive-org-7310 286 3 : : : web-archive-org-7310 286 4 Clustered clustered JJ web-archive-org-7310 286 5 rates rate NNS web-archive-org-7310 286 6 of of IN web-archive-org-7310 286 7 change change NN web-archive-org-7310 286 8 , , , web-archive-org-7310 286 9 broken break VBN web-archive-org-7310 286 10 down down RP web-archive-org-7310 286 11 by by IN web-archive-org-7310 286 12 selected select VBN web-archive-org-7310 286 13 top top JJ web-archive-org-7310 286 14 - - HYPH web-archive-org-7310 286 15 level level NN web-archive-org-7310 286 16 domains domain NNS web-archive-org-7310 286 17 , , , web-archive-org-7310 286 18 and and CC web-archive-org-7310 286 19 omitting omit VBG web-archive-org-7310 286 20 the the DT web-archive-org-7310 286 21 no no DT web-archive-org-7310 286 22 change change NN web-archive-org-7310 286 23 cluster cluster NN web-archive-org-7310 286 24 . . . web-archive-org-7310 287 1 Figure figure NN web-archive-org-7310 287 2   _SP web-archive-org-7310 287 3 9 9 CD web-archive-org-7310 287 4 is be VBZ web-archive-org-7310 287 5 similar similar JJ web-archive-org-7310 287 6 to to IN web-archive-org-7310 287 7 Figure figure NN web-archive-org-7310 287 8   _SP web-archive-org-7310 287 9 8 8 CD web-archive-org-7310 287 10 , , , web-archive-org-7310 287 11 but but CC web-archive-org-7310 287 12 excludes exclude VBZ web-archive-org-7310 287 13 the the DT web-archive-org-7310 287 14 same same JJ web-archive-org-7310 287 15 URLs url NNS web-archive-org-7310 287 16 that that WDT web-archive-org-7310 287 17 were be VBD web-archive-org-7310 287 18 excluded exclude VBN web-archive-org-7310 287 19 in in IN web-archive-org-7310 287 20 Figures figure NNS web-archive-org-7310 287 21   _SP web-archive-org-7310 287 22 11 11 CD web-archive-org-7310 287 23   _SP web-archive-org-7310 287 24 and and CC web-archive-org-7310 287 25   _SP web-archive-org-7310 287 26 12 12 CD web-archive-org-7310 287 27 . . . web-archive-org-7310 288 1 Note note VB web-archive-org-7310 288 2 that that IN web-archive-org-7310 288 3 most most JJS web-archive-org-7310 288 4 of of IN web-archive-org-7310 288 5 the the DT web-archive-org-7310 288 6 non non JJ web-archive-org-7310 288 7 - - JJ web-archive-org-7310 288 8 monotonicity monotonicity NN web-archive-org-7310 288 9 at at IN web-archive-org-7310 288 10 the the DT web-archive-org-7310 288 11 right right JJ web-archive-org-7310 288 12 end end NN web-archive-org-7310 288 13 of of IN web-archive-org-7310 288 14 the the DT web-archive-org-7310 288 15 distribution distribution NN web-archive-org-7310 288 16 has have VBZ web-archive-org-7310 288 17 disappeared disappear VBN web-archive-org-7310 288 18 , , , web-archive-org-7310 288 19 except except IN web-archive-org-7310 288 20 for for IN web-archive-org-7310 288 21 bucket bucket NN web-archive-org-7310 288 22 0 0 CD web-archive-org-7310 288 23 , , , web-archive-org-7310 288 24 which which WDT web-archive-org-7310 288 25 nonetheless nonetheless RB web-archive-org-7310 288 26 has have VBZ web-archive-org-7310 288 27 been be VBN web-archive-org-7310 288 28 cut cut VBN web-archive-org-7310 288 29 in in IN web-archive-org-7310 288 30 half half NN web-archive-org-7310 288 31 . . . web-archive-org-7310 289 1 We -PRON- PRP web-archive-org-7310 289 2 next next RB web-archive-org-7310 289 3 consider consider VBP web-archive-org-7310 289 4 whether whether IN web-archive-org-7310 289 5 the the DT web-archive-org-7310 289 6 length length NN web-archive-org-7310 289 7 of of IN web-archive-org-7310 289 8 pages page NNS web-archive-org-7310 289 9 impacts impact VBZ web-archive-org-7310 289 10 their -PRON- PRP$ web-archive-org-7310 289 11 rate rate NN web-archive-org-7310 289 12 of of IN web-archive-org-7310 289 13 change change NN web-archive-org-7310 289 14 . . . web-archive-org-7310 290 1 In in IN web-archive-org-7310 290 2 Figure figure NN web-archive-org-7310 290 3   _SP web-archive-org-7310 290 4 13 13 CD web-archive-org-7310 290 5 , , , web-archive-org-7310 290 6 we -PRON- PRP web-archive-org-7310 290 7 use use VBP web-archive-org-7310 290 8 the the DT web-archive-org-7310 290 9 same same JJ web-archive-org-7310 290 10 x x NN web-archive-org-7310 290 11 - - HYPH web-archive-org-7310 290 12 axis axis JJ web-archive-org-7310 290 13 semantics semantic NNS web-archive-org-7310 290 14 as as IN web-archive-org-7310 290 15 in in IN web-archive-org-7310 290 16 Figure figure NN web-archive-org-7310 290 17   _SP web-archive-org-7310 290 18 2 2 CD web-archive-org-7310 290 19 , , , web-archive-org-7310 290 20 and and CC web-archive-org-7310 290 21 the the DT web-archive-org-7310 290 22 same same JJ web-archive-org-7310 290 23 y y NN web-archive-org-7310 290 24 - - HYPH web-archive-org-7310 290 25 axis axis JJ web-archive-org-7310 290 26 semantics semantic NNS web-archive-org-7310 290 27 and and CC web-archive-org-7310 290 28 bar bar NN web-archive-org-7310 290 29 graph graph NN web-archive-org-7310 290 30 encodings encoding NNS web-archive-org-7310 290 31 as as IN web-archive-org-7310 290 32 in in IN web-archive-org-7310 290 33 Figure figure NN web-archive-org-7310 290 34   _SP web-archive-org-7310 290 35 10 10 CD web-archive-org-7310 290 36 . . . web-archive-org-7310 291 1 The the DT web-archive-org-7310 291 2 most most RBS web-archive-org-7310 291 3 striking striking JJ web-archive-org-7310 291 4 feature feature NN web-archive-org-7310 291 5 of of IN web-archive-org-7310 291 6 this this DT web-archive-org-7310 291 7 figure figure NN web-archive-org-7310 291 8 is be VBZ web-archive-org-7310 291 9 that that IN web-archive-org-7310 291 10 document document NN web-archive-org-7310 291 11 size size NN web-archive-org-7310 291 12 is be VBZ web-archive-org-7310 291 13 strongly strongly RB web-archive-org-7310 291 14 related relate VBN web-archive-org-7310 291 15 to to IN web-archive-org-7310 291 16 amount amount NN web-archive-org-7310 291 17 and and CC web-archive-org-7310 291 18 rate rate NN web-archive-org-7310 291 19 of of IN web-archive-org-7310 291 20 change change NN web-archive-org-7310 291 21 , , , web-archive-org-7310 291 22 and and CC web-archive-org-7310 291 23 counterintuitively counterintuitively RB web-archive-org-7310 291 24 so so RB web-archive-org-7310 291 25 ! ! . web-archive-org-7310 292 1 One one PRP web-archive-org-7310 292 2 might may MD web-archive-org-7310 292 3 think think VB web-archive-org-7310 292 4 that that IN web-archive-org-7310 292 5 small small JJ web-archive-org-7310 292 6 documents document NNS web-archive-org-7310 292 7 are be VBP web-archive-org-7310 292 8 more more RBR web-archive-org-7310 292 9 likely likely JJ web-archive-org-7310 292 10 to to TO web-archive-org-7310 292 11 change change VB web-archive-org-7310 292 12 , , , web-archive-org-7310 292 13 and and CC web-archive-org-7310 292 14 if if IN web-archive-org-7310 292 15 they -PRON- PRP web-archive-org-7310 292 16 do do VBP web-archive-org-7310 292 17 , , , web-archive-org-7310 292 18 change change VB web-archive-org-7310 292 19 more more RBR web-archive-org-7310 292 20 severely severely RB web-archive-org-7310 292 21 ( ( -LRB- web-archive-org-7310 292 22 since since IN web-archive-org-7310 292 23 any any DT web-archive-org-7310 292 24 change change NN web-archive-org-7310 292 25 is be VBZ web-archive-org-7310 292 26 a a DT web-archive-org-7310 292 27 large large JJ web-archive-org-7310 292 28 change change NN web-archive-org-7310 292 29 ) ) -RRB- web-archive-org-7310 292 30 . . . web-archive-org-7310 293 1 However however RB web-archive-org-7310 293 2 , , , web-archive-org-7310 293 3 we -PRON- PRP web-archive-org-7310 293 4 found find VBD web-archive-org-7310 293 5 that that IN web-archive-org-7310 293 6 large large JJ web-archive-org-7310 293 7 documents document NNS web-archive-org-7310 293 8 ( ( -LRB- web-archive-org-7310 293 9 32 32 CD web-archive-org-7310 293 10 KB KB NNP web-archive-org-7310 293 11 and and CC web-archive-org-7310 293 12 above above RB web-archive-org-7310 293 13 ) ) -RRB- web-archive-org-7310 293 14 change change VB web-archive-org-7310 293 15 much much RB web-archive-org-7310 293 16 more more RBR web-archive-org-7310 293 17 frequently frequently RB web-archive-org-7310 293 18 than than IN web-archive-org-7310 293 19 smaller small JJR web-archive-org-7310 293 20 ones one NNS web-archive-org-7310 293 21 ( ( -LRB- web-archive-org-7310 293 22 4 4 CD web-archive-org-7310 293 23 KB KB NNP web-archive-org-7310 293 24 and and CC web-archive-org-7310 293 25 below below RB web-archive-org-7310 293 26 ) ) -RRB- web-archive-org-7310 293 27 . . . web-archive-org-7310 294 1 Figure figure NN web-archive-org-7310 294 2 13 13 CD web-archive-org-7310 294 3 : : : web-archive-org-7310 294 4 Clustered clustered JJ web-archive-org-7310 294 5 rates rate NNS web-archive-org-7310 294 6 of of IN web-archive-org-7310 294 7 change change NN web-archive-org-7310 294 8 , , , web-archive-org-7310 294 9 broken break VBN web-archive-org-7310 294 10 down down RP web-archive-org-7310 294 11 by by IN web-archive-org-7310 294 12 document document NN web-archive-org-7310 294 13 size size NN web-archive-org-7310 294 14 . . . web-archive-org-7310 295 1 Figure figure NN web-archive-org-7310 295 2   _SP web-archive-org-7310 295 3 14 14 CD web-archive-org-7310 295 4 is be VBZ web-archive-org-7310 295 5 similar similar JJ web-archive-org-7310 295 6 in in IN web-archive-org-7310 295 7 spirit spirit NN web-archive-org-7310 295 8 , , , web-archive-org-7310 295 9 but but CC web-archive-org-7310 295 10 examines examine VBZ web-archive-org-7310 295 11 the the DT web-archive-org-7310 295 12 relationship relationship NN web-archive-org-7310 295 13 between between IN web-archive-org-7310 295 14 the the DT web-archive-org-7310 295 15 number number NN web-archive-org-7310 295 16 of of IN web-archive-org-7310 295 17 words word NNS web-archive-org-7310 295 18 and and CC web-archive-org-7310 295 19 the the DT web-archive-org-7310 295 20 rate rate NN web-archive-org-7310 295 21 of of IN web-archive-org-7310 295 22 change change NN web-archive-org-7310 295 23 . . . web-archive-org-7310 296 1 In in IN web-archive-org-7310 296 2 a a DT web-archive-org-7310 296 3 way way NN web-archive-org-7310 296 4 , , , web-archive-org-7310 296 5 this this DT web-archive-org-7310 296 6 metric metric NN web-archive-org-7310 296 7 is be VBZ web-archive-org-7310 296 8 more more RBR web-archive-org-7310 296 9 straightforward straightforward JJ web-archive-org-7310 296 10 , , , web-archive-org-7310 296 11 since since IN web-archive-org-7310 296 12 the the DT web-archive-org-7310 296 13 sensitivity sensitivity NN web-archive-org-7310 296 14 of of IN web-archive-org-7310 296 15 our -PRON- PRP$ web-archive-org-7310 296 16 shingling shingling NN web-archive-org-7310 296 17 techniques technique NNS web-archive-org-7310 296 18 depends depend VBZ web-archive-org-7310 296 19 on on IN web-archive-org-7310 296 20 the the DT web-archive-org-7310 296 21 number number NN web-archive-org-7310 296 22 of of IN web-archive-org-7310 296 23 words word NNS web-archive-org-7310 296 24 in in IN web-archive-org-7310 296 25 a a DT web-archive-org-7310 296 26 document document NN web-archive-org-7310 296 27 . . . web-archive-org-7310 297 1 For for IN web-archive-org-7310 297 2 documents document NNS web-archive-org-7310 297 3 with with IN web-archive-org-7310 297 4 few few JJ web-archive-org-7310 297 5 words word NNS web-archive-org-7310 297 6 , , , web-archive-org-7310 297 7 our -PRON- PRP$ web-archive-org-7310 297 8 metric metric JJ web-archive-org-7310 297 9 gives give VBZ web-archive-org-7310 297 10 a a DT web-archive-org-7310 297 11 relatively relatively RB web-archive-org-7310 297 12 coarse coarse JJ web-archive-org-7310 297 13 , , , web-archive-org-7310 297 14 " " `` web-archive-org-7310 297 15 all all DT web-archive-org-7310 297 16 - - HYPH web-archive-org-7310 297 17 or or CC web-archive-org-7310 297 18 - - HYPH web-archive-org-7310 297 19 nothing nothing NN web-archive-org-7310 297 20 " " '' web-archive-org-7310 297 21 similarity similarity NN web-archive-org-7310 297 22 metric metric NN web-archive-org-7310 297 23 . . . web-archive-org-7310 298 1 Nonetheless nonetheless RB web-archive-org-7310 298 2 , , , web-archive-org-7310 298 3 this this DT web-archive-org-7310 298 4 figure figure NN web-archive-org-7310 298 5 echoes echo VBZ web-archive-org-7310 298 6 the the DT web-archive-org-7310 298 7 observation observation NN web-archive-org-7310 298 8 of of IN web-archive-org-7310 298 9 Figure Figure NNP web-archive-org-7310 298 10   _SP web-archive-org-7310 298 11 13 13 CD web-archive-org-7310 298 12 , , , web-archive-org-7310 298 13 that that IN web-archive-org-7310 298 14 large large JJ web-archive-org-7310 298 15 documents document NNS web-archive-org-7310 298 16 are be VBP web-archive-org-7310 298 17 more more RBR web-archive-org-7310 298 18 likely likely JJ web-archive-org-7310 298 19 to to TO web-archive-org-7310 298 20 change change VB web-archive-org-7310 298 21 than than IN web-archive-org-7310 298 22 smaller small JJR web-archive-org-7310 298 23 ones one NNS web-archive-org-7310 298 24 . . . web-archive-org-7310 299 1 Figure figure NN web-archive-org-7310 299 2 14 14 CD web-archive-org-7310 299 3 : : : web-archive-org-7310 299 4 Clustered clustered JJ web-archive-org-7310 299 5 rates rate NNS web-archive-org-7310 299 6 of of IN web-archive-org-7310 299 7 change change NN web-archive-org-7310 299 8 , , , web-archive-org-7310 299 9 broken break VBN web-archive-org-7310 299 10 down down RP web-archive-org-7310 299 11 by by IN web-archive-org-7310 299 12 number number NN web-archive-org-7310 299 13 of of IN web-archive-org-7310 299 14 words word NNS web-archive-org-7310 299 15 per per IN web-archive-org-7310 299 16 document document NN web-archive-org-7310 299 17 . . . web-archive-org-7310 300 1 In in IN web-archive-org-7310 300 2 Figure figure NN web-archive-org-7310 300 3   _SP web-archive-org-7310 300 4 15 15 CD web-archive-org-7310 300 5 , , , web-archive-org-7310 300 6 we -PRON- PRP web-archive-org-7310 300 7 examine examine VBP web-archive-org-7310 300 8 the the DT web-archive-org-7310 300 9 same same JJ web-archive-org-7310 300 10 information information NN web-archive-org-7310 300 11 , , , web-archive-org-7310 300 12 excluding exclude VBG web-archive-org-7310 300 13 the the DT web-archive-org-7310 300 14 documents document NNS web-archive-org-7310 300 15 with with IN web-archive-org-7310 300 16 no no DT web-archive-org-7310 300 17 change change NN web-archive-org-7310 300 18 . . . web-archive-org-7310 301 1 We -PRON- PRP web-archive-org-7310 301 2 observe observe VBP web-archive-org-7310 301 3 that that IN web-archive-org-7310 301 4 pages page NNS web-archive-org-7310 301 5 with with IN web-archive-org-7310 301 6 different different JJ web-archive-org-7310 301 7 numbers number NNS web-archive-org-7310 301 8 of of IN web-archive-org-7310 301 9 words word NNS web-archive-org-7310 301 10 exhibit exhibit VBP web-archive-org-7310 301 11 similar similar JJ web-archive-org-7310 301 12 change change NN web-archive-org-7310 301 13 behavior behavior NN web-archive-org-7310 301 14 , , , web-archive-org-7310 301 15 except except IN web-archive-org-7310 301 16 that that IN web-archive-org-7310 301 17 pages page NNS web-archive-org-7310 301 18 with with IN web-archive-org-7310 301 19 just just RB web-archive-org-7310 301 20 a a DT web-archive-org-7310 301 21 few few JJ web-archive-org-7310 301 22 words word NNS web-archive-org-7310 301 23 can can MD web-archive-org-7310 301 24 not not RB web-archive-org-7310 301 25 show show VB web-archive-org-7310 301 26 an an DT web-archive-org-7310 301 27 intermediate intermediate JJ web-archive-org-7310 301 28 amount amount NN web-archive-org-7310 301 29 of of IN web-archive-org-7310 301 30 change change NN web-archive-org-7310 301 31 , , , web-archive-org-7310 301 32 due due IN web-archive-org-7310 301 33 to to IN web-archive-org-7310 301 34 our -PRON- PRP$ web-archive-org-7310 301 35 sampling sampling NN web-archive-org-7310 301 36 technique technique NN web-archive-org-7310 301 37 . . . web-archive-org-7310 302 1 Figure figure NN web-archive-org-7310 302 2 15 15 CD web-archive-org-7310 302 3 : : : web-archive-org-7310 302 4 Clustered clustered JJ web-archive-org-7310 302 5 rates rate NNS web-archive-org-7310 302 6 of of IN web-archive-org-7310 302 7 change change NN web-archive-org-7310 302 8 , , , web-archive-org-7310 302 9 broken break VBN web-archive-org-7310 302 10 down down RP web-archive-org-7310 302 11 by by IN web-archive-org-7310 302 12 number number NN web-archive-org-7310 302 13 of of IN web-archive-org-7310 302 14 words word NNS web-archive-org-7310 302 15 per per IN web-archive-org-7310 302 16 document document NN web-archive-org-7310 302 17 , , , web-archive-org-7310 302 18 and and CC web-archive-org-7310 302 19 omitting omit VBG web-archive-org-7310 302 20 the the DT web-archive-org-7310 302 21 no no DT web-archive-org-7310 302 22 change change NN web-archive-org-7310 302 23 cluster cluster NN web-archive-org-7310 302 24 . . . web-archive-org-7310 303 1 We -PRON- PRP web-archive-org-7310 303 2 further further RB web-archive-org-7310 303 3 investigated investigate VBD web-archive-org-7310 303 4 whether whether IN web-archive-org-7310 303 5 there there EX web-archive-org-7310 303 6 are be VBP web-archive-org-7310 303 7 any any DT web-archive-org-7310 303 8 confounding confound VBG web-archive-org-7310 303 9 relationships relationship NNS web-archive-org-7310 303 10 between between IN web-archive-org-7310 303 11 document document NN web-archive-org-7310 303 12 size size NN web-archive-org-7310 303 13 and and CC web-archive-org-7310 303 14 top top JJ web-archive-org-7310 303 15 - - HYPH web-archive-org-7310 303 16 level level NN web-archive-org-7310 303 17 domain domain NN web-archive-org-7310 303 18 . . . web-archive-org-7310 304 1 Figure figure NN web-archive-org-7310 304 2   _SP web-archive-org-7310 304 3 16 16 CD web-archive-org-7310 304 4 uses use VBZ web-archive-org-7310 304 5 the the DT web-archive-org-7310 304 6 same same JJ web-archive-org-7310 304 7 representation representation NN web-archive-org-7310 304 8 as as IN web-archive-org-7310 304 9 Figure figure NN web-archive-org-7310 304 10   _SP web-archive-org-7310 304 11 13 13 CD web-archive-org-7310 304 12 , , , web-archive-org-7310 304 13 but but CC web-archive-org-7310 304 14 each each DT web-archive-org-7310 304 15 chart chart NN web-archive-org-7310 304 16 considers consider VBZ web-archive-org-7310 304 17 only only RB web-archive-org-7310 304 18 those those DT web-archive-org-7310 304 19 URLs url NNS web-archive-org-7310 304 20 from from IN web-archive-org-7310 304 21 a a DT web-archive-org-7310 304 22 specific specific JJ web-archive-org-7310 304 23 top top JJ web-archive-org-7310 304 24 - - HYPH web-archive-org-7310 304 25 level level NN web-archive-org-7310 304 26 domain domain NN web-archive-org-7310 304 27 . . . web-archive-org-7310 305 1 The the DT web-archive-org-7310 305 2 distributions distribution NNS web-archive-org-7310 305 3 for for IN web-archive-org-7310 305 4 the the DT web-archive-org-7310 305 5 .com .com NN web-archive-org-7310 305 6 and and CC web-archive-org-7310 305 7 .net .net . web-archive-org-7310 305 8 domains domain NNS web-archive-org-7310 305 9 exhibit exhibit VBP web-archive-org-7310 305 10 a a DT web-archive-org-7310 305 11 much much RB web-archive-org-7310 305 12 stronger strong JJR web-archive-org-7310 305 13 threshold threshold NN web-archive-org-7310 305 14 effect effect NN web-archive-org-7310 305 15 for for IN web-archive-org-7310 305 16 large large JJ web-archive-org-7310 305 17 documents document NNS web-archive-org-7310 305 18 than than IN web-archive-org-7310 305 19 do do VBP web-archive-org-7310 305 20 .gov .gov . web-archive-org-7310 305 21 and and CC web-archive-org-7310 305 22 .edu .edu . web-archive-org-7310 305 23 . . . web-archive-org-7310 306 1 Figure figure NN web-archive-org-7310 306 2 16 16 CD web-archive-org-7310 306 3 : : : web-archive-org-7310 306 4 Clustered clustered JJ web-archive-org-7310 306 5 rates rate NNS web-archive-org-7310 306 6 of of IN web-archive-org-7310 306 7 change change NN web-archive-org-7310 306 8 , , , web-archive-org-7310 306 9 broken break VBN web-archive-org-7310 306 10 down down RP web-archive-org-7310 306 11 by by IN web-archive-org-7310 306 12 top top JJ web-archive-org-7310 306 13 - - HYPH web-archive-org-7310 306 14 level level NN web-archive-org-7310 306 15 domain domain NN web-archive-org-7310 306 16 and and CC web-archive-org-7310 306 17 number number NN web-archive-org-7310 306 18 of of IN web-archive-org-7310 306 19 words word NNS web-archive-org-7310 306 20 per per IN web-archive-org-7310 306 21 document document NN web-archive-org-7310 306 22 . . . web-archive-org-7310 307 1 Figure figure NN web-archive-org-7310 307 2   _SP web-archive-org-7310 307 3 17 17 CD web-archive-org-7310 307 4 examines examine NNS web-archive-org-7310 307 5 the the DT web-archive-org-7310 307 6 correlation correlation NN web-archive-org-7310 307 7 of of IN web-archive-org-7310 307 8 successive successive JJ web-archive-org-7310 307 9 changes change NNS web-archive-org-7310 307 10 to to IN web-archive-org-7310 307 11 a a DT web-archive-org-7310 307 12 document document NN web-archive-org-7310 307 13 . . . web-archive-org-7310 308 1 The the DT web-archive-org-7310 308 2 figure figure NN web-archive-org-7310 308 3 shows show VBZ web-archive-org-7310 308 4 a a DT web-archive-org-7310 308 5 3D 3d JJ web-archive-org-7310 308 6 histogram histogram NN web-archive-org-7310 308 7 . . . web-archive-org-7310 309 1 The the DT web-archive-org-7310 309 2 x x NN web-archive-org-7310 309 3 axis axis RB web-archive-org-7310 309 4 denotes denote VBZ web-archive-org-7310 309 5 the the DT web-archive-org-7310 309 6 number number NN web-archive-org-7310 309 7 of of IN web-archive-org-7310 309 8 pre pre NN web-archive-org-7310 309 9 - - NNS web-archive-org-7310 309 10 images image NNS web-archive-org-7310 309 11 in in IN web-archive-org-7310 309 12 a a DT web-archive-org-7310 309 13 document document NN web-archive-org-7310 309 14 unchanged unchanged JJ web-archive-org-7310 309 15 from from IN web-archive-org-7310 309 16 week week NN web-archive-org-7310 309 17 n-1 n-1 NN web-archive-org-7310 309 18 to to IN web-archive-org-7310 309 19 n n NNP web-archive-org-7310 309 20 , , , web-archive-org-7310 309 21 the the DT web-archive-org-7310 309 22 y y NN web-archive-org-7310 309 23 axis axis VBP web-archive-org-7310 309 24 the the DT web-archive-org-7310 309 25 number number NN web-archive-org-7310 309 26 of of IN web-archive-org-7310 309 27 pre pre JJ web-archive-org-7310 309 28 - - NNS web-archive-org-7310 309 29 images image NNS web-archive-org-7310 309 30 unchanged unchanged JJ web-archive-org-7310 309 31 from from IN web-archive-org-7310 309 32 week week NN web-archive-org-7310 309 33 n n IN web-archive-org-7310 309 34 to to IN web-archive-org-7310 309 35 n+1 n+1 NN web-archive-org-7310 309 36 , , , web-archive-org-7310 309 37 and and CC web-archive-org-7310 309 38 the the DT web-archive-org-7310 309 39 z z NN web-archive-org-7310 309 40 axis axis RB web-archive-org-7310 309 41 shows show VBZ web-archive-org-7310 309 42 the the DT web-archive-org-7310 309 43 logarithm logarithm NNP web-archive-org-7310 309 44 ( ( -LRB- web-archive-org-7310 309 45 base base NNP web-archive-org-7310 309 46 2 2 CD web-archive-org-7310 309 47 ) ) -RRB- web-archive-org-7310 309 48 of of IN web-archive-org-7310 309 49 the the DT web-archive-org-7310 309 50 number number NN web-archive-org-7310 309 51 of of IN web-archive-org-7310 309 52 such such JJ web-archive-org-7310 309 53 documents document NNS web-archive-org-7310 309 54 . . . web-archive-org-7310 310 1 A a DT web-archive-org-7310 310 2 data data NN web-archive-org-7310 310 3 point point NN web-archive-org-7310 310 4 ( ( -LRB- web-archive-org-7310 310 5 x x NNP web-archive-org-7310 310 6 , , , web-archive-org-7310 310 7 y y NNP web-archive-org-7310 310 8 , , , web-archive-org-7310 310 9 z z NNP web-archive-org-7310 310 10 ) ) -RRB- web-archive-org-7310 310 11 indicates indicate VBZ web-archive-org-7310 310 12 that that IN web-archive-org-7310 310 13 there there EX web-archive-org-7310 310 14 are be VBP web-archive-org-7310 310 15 2z 2z CD web-archive-org-7310 310 16 document document NN web-archive-org-7310 310 17 / / SYM web-archive-org-7310 310 18 week week NN web-archive-org-7310 310 19 pairs pair NNS web-archive-org-7310 310 20 ( ( -LRB- web-archive-org-7310 310 21 d d NNP web-archive-org-7310 310 22 , , , web-archive-org-7310 310 23 n n JJ web-archive-org-7310 310 24 ) ) -RRB- web-archive-org-7310 310 25 for for IN web-archive-org-7310 310 26 which which WDT web-archive-org-7310 310 27 the the DT web-archive-org-7310 310 28 versions version NNS web-archive-org-7310 310 29 of of IN web-archive-org-7310 310 30 document document NN web-archive-org-7310 310 31 d d NN web-archive-org-7310 310 32 had have VBD web-archive-org-7310 310 33 x x VBN web-archive-org-7310 310 34 pre pre VBN web-archive-org-7310 310 35 - - NNS web-archive-org-7310 310 36 images image NNS web-archive-org-7310 310 37 in in IN web-archive-org-7310 310 38 common common JJ web-archive-org-7310 310 39 between between IN web-archive-org-7310 310 40 weeks week NNS web-archive-org-7310 310 41 n-1 n-1 NNP web-archive-org-7310 310 42 and and CC web-archive-org-7310 310 43 n n NNP web-archive-org-7310 310 44 , , , web-archive-org-7310 310 45 and and CC web-archive-org-7310 310 46 y y NNP web-archive-org-7310 310 47 pre pre NN web-archive-org-7310 310 48 - - NNS web-archive-org-7310 310 49 images image NNS web-archive-org-7310 310 50 in in IN web-archive-org-7310 310 51 common common JJ web-archive-org-7310 310 52 between between IN web-archive-org-7310 310 53 weeks week NNS web-archive-org-7310 310 54 n n JJ web-archive-org-7310 310 55 and and CC web-archive-org-7310 310 56 n+1 n+1 JJ web-archive-org-7310 310 57 . . . web-archive-org-7310 311 1 The the DT web-archive-org-7310 311 2 spire spire NN web-archive-org-7310 311 3 surrounding surround VBG web-archive-org-7310 311 4 the the DT web-archive-org-7310 311 5 ( ( -LRB- web-archive-org-7310 311 6 x x NNP web-archive-org-7310 311 7 , , , web-archive-org-7310 311 8 y y NNP web-archive-org-7310 311 9 ) ) -RRB- web-archive-org-7310 311 10 coordinate coordinate NNP web-archive-org-7310 311 11 ( ( -LRB- web-archive-org-7310 311 12 85,85 85,85 CD web-archive-org-7310 311 13 ) ) -RRB- web-archive-org-7310 311 14 represents represent VBZ web-archive-org-7310 311 15 the the DT web-archive-org-7310 311 16 vast vast JJ web-archive-org-7310 311 17 majority majority NN web-archive-org-7310 311 18 of of IN web-archive-org-7310 311 19 web web NN web-archive-org-7310 311 20 pages page NNS web-archive-org-7310 311 21 that that WDT web-archive-org-7310 311 22 do do VBP web-archive-org-7310 311 23 n't not RB web-archive-org-7310 311 24 change change VB web-archive-org-7310 311 25 much much JJ web-archive-org-7310 311 26 over over IN web-archive-org-7310 311 27 a a DT web-archive-org-7310 311 28 three three CD web-archive-org-7310 311 29 - - HYPH web-archive-org-7310 311 30 week week NN web-archive-org-7310 311 31 interval interval NN web-archive-org-7310 311 32 . . . web-archive-org-7310 312 1 The the DT web-archive-org-7310 312 2 tip tip NN web-archive-org-7310 312 3 of of IN web-archive-org-7310 312 4 the the DT web-archive-org-7310 312 5 spire spire NN web-archive-org-7310 312 6 is be VBZ web-archive-org-7310 312 7 ten ten CD web-archive-org-7310 312 8 thousand thousand CD web-archive-org-7310 312 9 times time NNS web-archive-org-7310 312 10 higher high JJR web-archive-org-7310 312 11 than than IN web-archive-org-7310 312 12 any any DT web-archive-org-7310 312 13 other other JJ web-archive-org-7310 312 14 feature feature NN web-archive-org-7310 312 15 in in IN web-archive-org-7310 312 16 the the DT web-archive-org-7310 312 17 plot plot NN web-archive-org-7310 312 18 , , , web-archive-org-7310 312 19 except except IN web-archive-org-7310 312 20 for for IN web-archive-org-7310 312 21 the the DT web-archive-org-7310 312 22 smaller small JJR web-archive-org-7310 312 23 spire spire NN web-archive-org-7310 312 24 at at IN web-archive-org-7310 312 25 the the DT web-archive-org-7310 312 26 other other JJ web-archive-org-7310 312 27 end end NN web-archive-org-7310 312 28 of of IN web-archive-org-7310 312 29 the the DT web-archive-org-7310 312 30 diagonal diagonal JJ web-archive-org-7310 312 31 , , , web-archive-org-7310 312 32 which which WDT web-archive-org-7310 312 33 represents represent VBZ web-archive-org-7310 312 34 documents document NNS web-archive-org-7310 312 35 which which WDT web-archive-org-7310 312 36 differ differ VBP web-archive-org-7310 312 37 completely completely RB web-archive-org-7310 312 38 in in IN web-archive-org-7310 312 39 every every DT web-archive-org-7310 312 40 sample sample NN web-archive-org-7310 312 41 . . . web-archive-org-7310 313 1 Much much JJ web-archive-org-7310 313 2 of of IN web-archive-org-7310 313 3 this this DT web-archive-org-7310 313 4 second second JJ web-archive-org-7310 313 5 peak peak NN web-archive-org-7310 313 6 can can MD web-archive-org-7310 313 7 be be VB web-archive-org-7310 313 8 attributed attribute VBN web-archive-org-7310 313 9 to to IN web-archive-org-7310 313 10 machine machine NN web-archive-org-7310 313 11 - - HYPH web-archive-org-7310 313 12 generated generate VBN web-archive-org-7310 313 13 pornography pornography NN web-archive-org-7310 313 14 , , , web-archive-org-7310 313 15 as as IN web-archive-org-7310 313 16 described describe VBN web-archive-org-7310 313 17 above above RB web-archive-org-7310 313 18 . . . web-archive-org-7310 314 1 The the DT web-archive-org-7310 314 2 second second RB web-archive-org-7310 314 3 - - HYPH web-archive-org-7310 314 4 most most RBS web-archive-org-7310 314 5 prominent prominent JJ web-archive-org-7310 314 6 feature feature NN web-archive-org-7310 314 7 is be VBZ web-archive-org-7310 314 8 the the DT web-archive-org-7310 314 9 pronounced pronounced JJ web-archive-org-7310 314 10 ridge ridge NN web-archive-org-7310 314 11 along along IN web-archive-org-7310 314 12 the the DT web-archive-org-7310 314 13 main main JJ web-archive-org-7310 314 14 diagonal diagonal NN web-archive-org-7310 314 15 of of IN web-archive-org-7310 314 16 the the DT web-archive-org-7310 314 17 xy xy NNP web-archive-org-7310 314 18 plane plane NN web-archive-org-7310 314 19 . . . web-archive-org-7310 315 1 The the DT web-archive-org-7310 315 2 crest crest NN web-archive-org-7310 315 3 of of IN web-archive-org-7310 315 4 the the DT web-archive-org-7310 315 5 ridge ridge NN web-archive-org-7310 315 6 represents represent VBZ web-archive-org-7310 315 7 a a DT web-archive-org-7310 315 8 thousandfold thousandfold JJ web-archive-org-7310 315 9 higher high JJR web-archive-org-7310 315 10 number number NN web-archive-org-7310 315 11 of of IN web-archive-org-7310 315 12 instances instance NNS web-archive-org-7310 315 13 per per IN web-archive-org-7310 315 14 grid grid NN web-archive-org-7310 315 15 point point NN web-archive-org-7310 315 16 than than IN web-archive-org-7310 315 17 the the DT web-archive-org-7310 315 18 floor floor NN web-archive-org-7310 315 19 of of IN web-archive-org-7310 315 20 the the DT web-archive-org-7310 315 21 valley valley NN web-archive-org-7310 315 22 . . . web-archive-org-7310 316 1 This this DT web-archive-org-7310 316 2 ridge ridge NN web-archive-org-7310 316 3 suggests suggest VBZ web-archive-org-7310 316 4 that that IN web-archive-org-7310 316 5 changes change NNS web-archive-org-7310 316 6 are be VBP web-archive-org-7310 316 7 highly highly RB web-archive-org-7310 316 8 correlated correlate VBN web-archive-org-7310 316 9 ; ; : web-archive-org-7310 316 10 past past JJ web-archive-org-7310 316 11 changes change NNS web-archive-org-7310 316 12 to to IN web-archive-org-7310 316 13 a a DT web-archive-org-7310 316 14 document document NN web-archive-org-7310 316 15 are be VBP web-archive-org-7310 316 16 an an DT web-archive-org-7310 316 17 excellent excellent JJ web-archive-org-7310 316 18 predictor predictor NN web-archive-org-7310 316 19 of of IN web-archive-org-7310 316 20 future future JJ web-archive-org-7310 316 21 changes change NNS web-archive-org-7310 316 22 . . . web-archive-org-7310 317 1 The the DT web-archive-org-7310 317 2 plumes plume NNS web-archive-org-7310 317 3 at at IN web-archive-org-7310 317 4 the the DT web-archive-org-7310 317 5 far far JJ web-archive-org-7310 317 6 walls wall NNS web-archive-org-7310 317 7 of of IN web-archive-org-7310 317 8 the the DT web-archive-org-7310 317 9 plot plot NN web-archive-org-7310 317 10 demonstrate demonstrate NN web-archive-org-7310 317 11 that that IN web-archive-org-7310 317 12 a a DT web-archive-org-7310 317 13 sizable sizable JJ web-archive-org-7310 317 14 fraction fraction NN web-archive-org-7310 317 15 of of IN web-archive-org-7310 317 16 documents document NNS web-archive-org-7310 317 17 do do VBP web-archive-org-7310 317 18 n't not RB web-archive-org-7310 317 19 change change VB web-archive-org-7310 317 20 in in IN web-archive-org-7310 317 21 a a DT web-archive-org-7310 317 22 given give VBN web-archive-org-7310 317 23 week week NN web-archive-org-7310 317 24 , , , web-archive-org-7310 317 25 even even RB web-archive-org-7310 317 26 if if IN web-archive-org-7310 317 27 they -PRON- PRP web-archive-org-7310 317 28 changed change VBD web-archive-org-7310 317 29 in in IN web-archive-org-7310 317 30 the the DT web-archive-org-7310 317 31 previous previous JJ web-archive-org-7310 317 32 or or CC web-archive-org-7310 317 33 following follow VBG web-archive-org-7310 317 34 week week NN web-archive-org-7310 317 35 . . . web-archive-org-7310 318 1 Figure figure NN web-archive-org-7310 318 2 17 17 CD web-archive-org-7310 318 3 : : : web-archive-org-7310 318 4 Logarithmic Logarithmic NNP web-archive-org-7310 318 5 histogram histogram NNP web-archive-org-7310 318 6 of of IN web-archive-org-7310 318 7 intra intra JJ web-archive-org-7310 318 8 - - JJ web-archive-org-7310 318 9 document document JJ web-archive-org-7310 318 10 changes change NNS web-archive-org-7310 318 11 over over IN web-archive-org-7310 318 12 three three CD web-archive-org-7310 318 13 successive successive JJ web-archive-org-7310 318 14 weeks week NNS web-archive-org-7310 318 15 , , , web-archive-org-7310 318 16 showing show VBG web-archive-org-7310 318 17 absolute absolute JJ web-archive-org-7310 318 18 number number NN web-archive-org-7310 318 19 of of IN web-archive-org-7310 318 20 changes change NNS web-archive-org-7310 318 21 . . . web-archive-org-7310 319 1 Figure figure NN web-archive-org-7310 319 2   _SP web-archive-org-7310 319 3 18 18 CD web-archive-org-7310 319 4 modifies modifie NNS web-archive-org-7310 319 5 the the DT web-archive-org-7310 319 6 previous previous JJ web-archive-org-7310 319 7 figure figure NN web-archive-org-7310 319 8 in in IN web-archive-org-7310 319 9 two two CD web-archive-org-7310 319 10 ways way NNS web-archive-org-7310 319 11 : : : web-archive-org-7310 319 12 The the DT web-archive-org-7310 319 13 view view NN web-archive-org-7310 319 14 is be VBZ web-archive-org-7310 319 15 down down IN web-archive-org-7310 319 16 the the DT web-archive-org-7310 319 17 z z NN web-archive-org-7310 319 18 axis axi NNS web-archive-org-7310 319 19 , , , web-archive-org-7310 319 20 transforming transform VBG web-archive-org-7310 319 21 the the DT web-archive-org-7310 319 22 3D 3d JJ web-archive-org-7310 319 23 plot plot NN web-archive-org-7310 319 24 into into IN web-archive-org-7310 319 25 a a DT web-archive-org-7310 319 26 2D 2d JJ web-archive-org-7310 319 27 contour contour NN web-archive-org-7310 319 28 map map NN web-archive-org-7310 319 29 , , , web-archive-org-7310 319 30 where where WRB web-archive-org-7310 319 31 color color NN web-archive-org-7310 319 32 / / SYM web-archive-org-7310 319 33 shading shade VBG web-archive-org-7310 319 34 indicate indicate VBP web-archive-org-7310 319 35 the the DT web-archive-org-7310 319 36 elevation elevation NN web-archive-org-7310 319 37 of of IN web-archive-org-7310 319 38 the the DT web-archive-org-7310 319 39 terrain terrain NN web-archive-org-7310 319 40 . . . web-archive-org-7310 320 1 In in IN web-archive-org-7310 320 2 addition addition NN web-archive-org-7310 320 3 , , , web-archive-org-7310 320 4 rather rather RB web-archive-org-7310 320 5 than than IN web-archive-org-7310 320 6 displaying display VBG web-archive-org-7310 320 7 absolute absolute JJ web-archive-org-7310 320 8 numbers number NNS web-archive-org-7310 320 9 of of IN web-archive-org-7310 320 10 samples sample NNS web-archive-org-7310 320 11 , , , web-archive-org-7310 320 12 we -PRON- PRP web-archive-org-7310 320 13 consider consider VBP web-archive-org-7310 320 14 each each DT web-archive-org-7310 320 15 column column NN web-archive-org-7310 320 16 as as IN web-archive-org-7310 320 17 a a DT web-archive-org-7310 320 18 probability probability NN web-archive-org-7310 320 19 distribution distribution NN web-archive-org-7310 320 20 ( ( -LRB- web-archive-org-7310 320 21 meaning mean VBG web-archive-org-7310 320 22 that that IN web-archive-org-7310 320 23 every every DT web-archive-org-7310 320 24 data data NN web-archive-org-7310 320 25 point point NN web-archive-org-7310 320 26 is be VBZ web-archive-org-7310 320 27 divided divide VBN web-archive-org-7310 320 28 by by IN web-archive-org-7310 320 29 the the DT web-archive-org-7310 320 30 sum sum NN web-archive-org-7310 320 31 of of IN web-archive-org-7310 320 32 the the DT web-archive-org-7310 320 33 data data NN web-archive-org-7310 320 34 points point NNS web-archive-org-7310 320 35 in in IN web-archive-org-7310 320 36 its -PRON- PRP$ web-archive-org-7310 320 37 column column NN web-archive-org-7310 320 38 ) ) -RRB- web-archive-org-7310 320 39 . . . web-archive-org-7310 321 1 Since since IN web-archive-org-7310 321 2 these these DT web-archive-org-7310 321 3 values value NNS web-archive-org-7310 321 4 range range VBP web-archive-org-7310 321 5 from from IN web-archive-org-7310 321 6 0 0 CD web-archive-org-7310 321 7 to to IN web-archive-org-7310 321 8 1 1 CD web-archive-org-7310 321 9 , , , web-archive-org-7310 321 10 their -PRON- PRP$ web-archive-org-7310 321 11 logarithms logarithm NNS web-archive-org-7310 321 12 are be VBP web-archive-org-7310 321 13 negative negative JJ web-archive-org-7310 321 14 . . . web-archive-org-7310 322 1 This this DT web-archive-org-7310 322 2 normalization normalization NN web-archive-org-7310 322 3 eliminates eliminate VBZ web-archive-org-7310 322 4 the the DT web-archive-org-7310 322 5 spires spire NNS web-archive-org-7310 322 6 that that WDT web-archive-org-7310 322 7 were be VBD web-archive-org-7310 322 8 so so RB web-archive-org-7310 322 9 prominent prominent JJ web-archive-org-7310 322 10 in in IN web-archive-org-7310 322 11 the the DT web-archive-org-7310 322 12 previous previous JJ web-archive-org-7310 322 13 figure figure NN web-archive-org-7310 322 14 . . . web-archive-org-7310 323 1 The the DT web-archive-org-7310 323 2 diagonal diagonal JJ web-archive-org-7310 323 3 ridge ridge NN web-archive-org-7310 323 4 , , , web-archive-org-7310 323 5 however however RB web-archive-org-7310 323 6 , , , web-archive-org-7310 323 7 remains remain VBZ web-archive-org-7310 323 8 , , , web-archive-org-7310 323 9 indicating indicate VBG web-archive-org-7310 323 10 once once RB web-archive-org-7310 323 11 again again RB web-archive-org-7310 323 12 that that IN web-archive-org-7310 323 13 past past JJ web-archive-org-7310 323 14 change change NN web-archive-org-7310 323 15 is be VBZ web-archive-org-7310 323 16 a a DT web-archive-org-7310 323 17 strong strong JJ web-archive-org-7310 323 18 predictor predictor NN web-archive-org-7310 323 19 of of IN web-archive-org-7310 323 20 future future JJ web-archive-org-7310 323 21 change change NN web-archive-org-7310 323 22 . . . web-archive-org-7310 324 1 Likewise likewise RB web-archive-org-7310 324 2 , , , web-archive-org-7310 324 3 the the DT web-archive-org-7310 324 4 plume plume NN web-archive-org-7310 324 5 along along IN web-archive-org-7310 324 6 the the DT web-archive-org-7310 324 7 top top NN web-archive-org-7310 324 8 remains remain VBZ web-archive-org-7310 324 9 clearly clearly RB web-archive-org-7310 324 10 visible visible JJ web-archive-org-7310 324 11 . . . web-archive-org-7310 325 1 Figure figure NN web-archive-org-7310 325 2 18 18 CD web-archive-org-7310 325 3 : : : web-archive-org-7310 325 4 Logarithmic Logarithmic NNP web-archive-org-7310 325 5 histogram histogram NNP web-archive-org-7310 325 6 of of IN web-archive-org-7310 325 7 intra intra JJ web-archive-org-7310 325 8 - - JJ web-archive-org-7310 325 9 document document JJ web-archive-org-7310 325 10 changes change NNS web-archive-org-7310 325 11 over over IN web-archive-org-7310 325 12 three three CD web-archive-org-7310 325 13 successive successive JJ web-archive-org-7310 325 14 weeks week NNS web-archive-org-7310 325 15 , , , web-archive-org-7310 325 16 normalized normalize VBD web-archive-org-7310 325 17 to to TO web-archive-org-7310 325 18 show show VB web-archive-org-7310 325 19 conditional conditional JJ web-archive-org-7310 325 20 probabilities probability NNS web-archive-org-7310 325 21 of of IN web-archive-org-7310 325 22 changes change NNS web-archive-org-7310 325 23 . . . web-archive-org-7310 326 1 5 5 CD web-archive-org-7310 326 2 . . . web-archive-org-7310 327 1 CONCLUSIONS conclusion NNS web-archive-org-7310 327 2 This this DT web-archive-org-7310 327 3 paper paper NN web-archive-org-7310 327 4 describes describe VBZ web-archive-org-7310 327 5 a a DT web-archive-org-7310 327 6 large large JJ web-archive-org-7310 327 7 - - HYPH web-archive-org-7310 327 8 scale scale NN web-archive-org-7310 327 9 experiment experiment NN web-archive-org-7310 327 10 aimed aim VBN web-archive-org-7310 327 11 at at IN web-archive-org-7310 327 12 measuring measure VBG web-archive-org-7310 327 13 the the DT web-archive-org-7310 327 14 rate rate NN web-archive-org-7310 327 15 and and CC web-archive-org-7310 327 16 degree degree NN web-archive-org-7310 327 17 of of IN web-archive-org-7310 327 18 web web NN web-archive-org-7310 327 19 page page NN web-archive-org-7310 327 20 changes change NNS web-archive-org-7310 327 21 over over IN web-archive-org-7310 327 22 a a DT web-archive-org-7310 327 23 significant significant JJ web-archive-org-7310 327 24 period period NN web-archive-org-7310 327 25 of of IN web-archive-org-7310 327 26 time time NN web-archive-org-7310 327 27 . . . web-archive-org-7310 328 1 We -PRON- PRP web-archive-org-7310 328 2 crawled crawl VBD web-archive-org-7310 328 3 151 151 CD web-archive-org-7310 328 4 million million CD web-archive-org-7310 328 5 pages page NNS web-archive-org-7310 328 6 once once RB web-archive-org-7310 328 7 a a DT web-archive-org-7310 328 8 week week NN web-archive-org-7310 328 9 for for IN web-archive-org-7310 328 10 eleven eleven CD web-archive-org-7310 328 11 weeks week NNS web-archive-org-7310 328 12 , , , web-archive-org-7310 328 13 saving save VBG web-archive-org-7310 328 14 salient salient NN web-archive-org-7310 328 15 information information NN web-archive-org-7310 328 16 about about IN web-archive-org-7310 328 17 each each DT web-archive-org-7310 328 18 downloaded download VBN web-archive-org-7310 328 19 document document NN web-archive-org-7310 328 20 , , , web-archive-org-7310 328 21 including include VBG web-archive-org-7310 328 22 a a DT web-archive-org-7310 328 23 feature feature NN web-archive-org-7310 328 24 vector vector NN web-archive-org-7310 328 25 of of IN web-archive-org-7310 328 26 the the DT web-archive-org-7310 328 27 text text NN web-archive-org-7310 328 28 without without IN web-archive-org-7310 328 29 markup markup NNP web-archive-org-7310 328 30 , , , web-archive-org-7310 328 31 plus plus CC web-archive-org-7310 328 32 the the DT web-archive-org-7310 328 33 full full JJ web-archive-org-7310 328 34 text text NN web-archive-org-7310 328 35 of of IN web-archive-org-7310 328 36 0.1 0.1 CD web-archive-org-7310 328 37 % % NN web-archive-org-7310 328 38 of of IN web-archive-org-7310 328 39 all all DT web-archive-org-7310 328 40 downloaded download VBN web-archive-org-7310 328 41 pages page NNS web-archive-org-7310 328 42 . . . web-archive-org-7310 329 1 Subsequently subsequently RB web-archive-org-7310 329 2 , , , web-archive-org-7310 329 3 we -PRON- PRP web-archive-org-7310 329 4 distilled distil VBD web-archive-org-7310 329 5 the the DT web-archive-org-7310 329 6 retained retain VBN web-archive-org-7310 329 7 data datum NNS web-archive-org-7310 329 8 to to TO web-archive-org-7310 329 9 make make VB web-archive-org-7310 329 10 it -PRON- PRP web-archive-org-7310 329 11 more more RBR web-archive-org-7310 329 12 amenable amenable JJ web-archive-org-7310 329 13 to to IN web-archive-org-7310 329 14 statistical statistical JJ web-archive-org-7310 329 15 analysis analysis NN web-archive-org-7310 329 16 , , , web-archive-org-7310 329 17 and and CC web-archive-org-7310 329 18 we -PRON- PRP web-archive-org-7310 329 19 performed perform VBD web-archive-org-7310 329 20 a a DT web-archive-org-7310 329 21 number number NN web-archive-org-7310 329 22 of of IN web-archive-org-7310 329 23 data datum NNS web-archive-org-7310 329 24 mining mining NN web-archive-org-7310 329 25 operations operation NNS web-archive-org-7310 329 26 on on IN web-archive-org-7310 329 27 the the DT web-archive-org-7310 329 28 distilled distil VBN web-archive-org-7310 329 29 data datum NNS web-archive-org-7310 329 30 . . . web-archive-org-7310 330 1 We -PRON- PRP web-archive-org-7310 330 2 found find VBD web-archive-org-7310 330 3 that that IN web-archive-org-7310 330 4 web web NN web-archive-org-7310 330 5 pages page NNS web-archive-org-7310 330 6 that that WDT web-archive-org-7310 330 7 change change VBP web-archive-org-7310 330 8 usually usually RB web-archive-org-7310 330 9 either either DT web-archive-org-7310 330 10 change change NN web-archive-org-7310 330 11 only only RB web-archive-org-7310 330 12 in in IN web-archive-org-7310 330 13 their -PRON- PRP$ web-archive-org-7310 330 14 markup markup NN web-archive-org-7310 330 15 or or CC web-archive-org-7310 330 16 in in IN web-archive-org-7310 330 17 trivial trivial JJ web-archive-org-7310 330 18 ways way NNS web-archive-org-7310 330 19 . . . web-archive-org-7310 331 1 Moreover moreover RB web-archive-org-7310 331 2 , , , web-archive-org-7310 331 3 we -PRON- PRP web-archive-org-7310 331 4 found find VBD web-archive-org-7310 331 5 that that IN web-archive-org-7310 331 6 there there EX web-archive-org-7310 331 7 is be VBZ web-archive-org-7310 331 8 a a DT web-archive-org-7310 331 9 strong strong JJ web-archive-org-7310 331 10 relationship relationship NN web-archive-org-7310 331 11 between between IN web-archive-org-7310 331 12 the the DT web-archive-org-7310 331 13 top top JJ web-archive-org-7310 331 14 - - HYPH web-archive-org-7310 331 15 level level NN web-archive-org-7310 331 16 domain domain NN web-archive-org-7310 331 17 and and CC web-archive-org-7310 331 18 the the DT web-archive-org-7310 331 19 frequency frequency NN web-archive-org-7310 331 20 of of IN web-archive-org-7310 331 21 change change NN web-archive-org-7310 331 22 of of IN web-archive-org-7310 331 23 a a DT web-archive-org-7310 331 24 document document NN web-archive-org-7310 331 25 , , , web-archive-org-7310 331 26 whereas whereas IN web-archive-org-7310 331 27 the the DT web-archive-org-7310 331 28 relationship relationship NN web-archive-org-7310 331 29 between between IN web-archive-org-7310 331 30 top top JJ web-archive-org-7310 331 31 - - HYPH web-archive-org-7310 331 32 level level NN web-archive-org-7310 331 33 domain domain NN web-archive-org-7310 331 34 and and CC web-archive-org-7310 331 35 degree degree NN web-archive-org-7310 331 36 of of IN web-archive-org-7310 331 37 change change NN web-archive-org-7310 331 38 is be VBZ web-archive-org-7310 331 39 much much RB web-archive-org-7310 331 40 weaker weak JJR web-archive-org-7310 331 41 . . . web-archive-org-7310 332 1 To to IN web-archive-org-7310 332 2 our -PRON- PRP$ web-archive-org-7310 332 3 great great JJ web-archive-org-7310 332 4 surprise surprise NN web-archive-org-7310 332 5 , , , web-archive-org-7310 332 6 we -PRON- PRP web-archive-org-7310 332 7 found find VBD web-archive-org-7310 332 8 that that IN web-archive-org-7310 332 9 document document NN web-archive-org-7310 332 10 size size NN web-archive-org-7310 332 11 is be VBZ web-archive-org-7310 332 12 another another DT web-archive-org-7310 332 13 strong strong JJ web-archive-org-7310 332 14 predictor predictor NN web-archive-org-7310 332 15 of of IN web-archive-org-7310 332 16 both both DT web-archive-org-7310 332 17 frequency frequency NN web-archive-org-7310 332 18 and and CC web-archive-org-7310 332 19 degree degree NN web-archive-org-7310 332 20 of of IN web-archive-org-7310 332 21 change change NN web-archive-org-7310 332 22 . . . web-archive-org-7310 333 1 Moreover moreover RB web-archive-org-7310 333 2 , , , web-archive-org-7310 333 3 one one PRP web-archive-org-7310 333 4 might may MD web-archive-org-7310 333 5 expect expect VB web-archive-org-7310 333 6 that that IN web-archive-org-7310 333 7 any any DT web-archive-org-7310 333 8 change change NN web-archive-org-7310 333 9 to to IN web-archive-org-7310 333 10 a a DT web-archive-org-7310 333 11 small small JJ web-archive-org-7310 333 12 document document NN web-archive-org-7310 333 13 would would MD web-archive-org-7310 333 14 be be VB web-archive-org-7310 333 15 a a DT web-archive-org-7310 333 16 significant significant JJ web-archive-org-7310 333 17 one one NN web-archive-org-7310 333 18 , , , web-archive-org-7310 333 19 by by IN web-archive-org-7310 333 20 virtue virtue NN web-archive-org-7310 333 21 of of IN web-archive-org-7310 333 22 small small JJ web-archive-org-7310 333 23 documents document NNS web-archive-org-7310 333 24 having have VBG web-archive-org-7310 333 25 fewer few JJR web-archive-org-7310 333 26 words word NNS web-archive-org-7310 333 27 , , , web-archive-org-7310 333 28 so so IN web-archive-org-7310 333 29 that that IN web-archive-org-7310 333 30 any any DT web-archive-org-7310 333 31 word word NN web-archive-org-7310 333 32 change change NN web-archive-org-7310 333 33 affects affect VBZ web-archive-org-7310 333 34 a a DT web-archive-org-7310 333 35 significant significant JJ web-archive-org-7310 333 36 fraction fraction NN web-archive-org-7310 333 37 of of IN web-archive-org-7310 333 38 the the DT web-archive-org-7310 333 39 shingles shingle NNS web-archive-org-7310 333 40 . . . web-archive-org-7310 334 1 Contrary contrary JJ web-archive-org-7310 334 2 to to IN web-archive-org-7310 334 3 that that DT web-archive-org-7310 334 4 intuition intuition NN web-archive-org-7310 334 5 , , , web-archive-org-7310 334 6 we -PRON- PRP web-archive-org-7310 334 7 found find VBD web-archive-org-7310 334 8 that that IN web-archive-org-7310 334 9 large large JJ web-archive-org-7310 334 10 documents document NNS web-archive-org-7310 334 11 change change VBP web-archive-org-7310 334 12 more more RBR web-archive-org-7310 334 13 often often RB web-archive-org-7310 334 14 and and CC web-archive-org-7310 334 15 more more RBR web-archive-org-7310 334 16 extensively extensively RB web-archive-org-7310 334 17 than than IN web-archive-org-7310 334 18 smaller small JJR web-archive-org-7310 334 19 ones one NNS web-archive-org-7310 334 20 . . . web-archive-org-7310 335 1 We -PRON- PRP web-archive-org-7310 335 2 investigated investigate VBD web-archive-org-7310 335 3 whether whether IN web-archive-org-7310 335 4 the the DT web-archive-org-7310 335 5 two two CD web-archive-org-7310 335 6 factors factor NNS web-archive-org-7310 335 7 - - HYPH web-archive-org-7310 335 8 top top JJ web-archive-org-7310 335 9 - - HYPH web-archive-org-7310 335 10 level level NN web-archive-org-7310 335 11 domain domain NN web-archive-org-7310 335 12 and and CC web-archive-org-7310 335 13 document document NN web-archive-org-7310 335 14 size size NN web-archive-org-7310 335 15 - - , web-archive-org-7310 335 16 were be VBD web-archive-org-7310 335 17 confounding confound VBG web-archive-org-7310 335 18 , , , web-archive-org-7310 335 19 and and CC web-archive-org-7310 335 20 discovered discover VBD web-archive-org-7310 335 21 that that IN web-archive-org-7310 335 22 for for IN web-archive-org-7310 335 23 the the DT web-archive-org-7310 335 24 most most JJS web-archive-org-7310 335 25 part part NN web-archive-org-7310 335 26 , , , web-archive-org-7310 335 27 the the DT web-archive-org-7310 335 28 relationship relationship NN web-archive-org-7310 335 29 of of IN web-archive-org-7310 335 30 document document NN web-archive-org-7310 335 31 size size NN web-archive-org-7310 335 32 and and CC web-archive-org-7310 335 33 rate rate NN web-archive-org-7310 335 34 and and CC web-archive-org-7310 335 35 degree degree NN web-archive-org-7310 335 36 of of IN web-archive-org-7310 335 37 change change NN web-archive-org-7310 335 38 are be VBP web-archive-org-7310 335 39 more more RBR web-archive-org-7310 335 40 pronounced pronounced JJ web-archive-org-7310 335 41 for for IN web-archive-org-7310 335 42 the the DT web-archive-org-7310 335 43 .com .com NFP web-archive-org-7310 335 44 and and CC web-archive-org-7310 335 45 .net .net . web-archive-org-7310 335 46 domains domain NNS web-archive-org-7310 335 47 than than IN web-archive-org-7310 335 48 for for IN web-archive-org-7310 335 49 , , , web-archive-org-7310 335 50 say say UH web-archive-org-7310 335 51 , , , web-archive-org-7310 335 52 the the DT web-archive-org-7310 335 53 .edu .edu NFP web-archive-org-7310 335 54 and and CC web-archive-org-7310 335 55 .gov .gov CD web-archive-org-7310 335 56 domains domain NNS web-archive-org-7310 335 57 , , , web-archive-org-7310 335 58 suggesting suggest VBG web-archive-org-7310 335 59 that that IN web-archive-org-7310 335 60 they -PRON- PRP web-archive-org-7310 335 61 are be VBP web-archive-org-7310 335 62 not not RB web-archive-org-7310 335 63 confounded confound VBN web-archive-org-7310 335 64 . . . web-archive-org-7310 336 1 We -PRON- PRP web-archive-org-7310 336 2 also also RB web-archive-org-7310 336 3 found find VBD web-archive-org-7310 336 4 that that IN web-archive-org-7310 336 5 past past JJ web-archive-org-7310 336 6 changes change NNS web-archive-org-7310 336 7 to to IN web-archive-org-7310 336 8 a a DT web-archive-org-7310 336 9 page page NN web-archive-org-7310 336 10 are be VBP web-archive-org-7310 336 11 a a DT web-archive-org-7310 336 12 good good JJ web-archive-org-7310 336 13 predictor predictor NN web-archive-org-7310 336 14 of of IN web-archive-org-7310 336 15 future future JJ web-archive-org-7310 336 16 changes change NNS web-archive-org-7310 336 17 . . . web-archive-org-7310 337 1 This this DT web-archive-org-7310 337 2 result result NN web-archive-org-7310 337 3 has have VBZ web-archive-org-7310 337 4 practical practical JJ web-archive-org-7310 337 5 implications implication NNS web-archive-org-7310 337 6 for for IN web-archive-org-7310 337 7 incremental incremental JJ web-archive-org-7310 337 8 web web NN web-archive-org-7310 337 9 crawlers crawler NNS web-archive-org-7310 337 10 that that WDT web-archive-org-7310 337 11 seek seek VBP web-archive-org-7310 337 12 to to TO web-archive-org-7310 337 13 maximize maximize VB web-archive-org-7310 337 14 the the DT web-archive-org-7310 337 15 freshness freshness NN web-archive-org-7310 337 16 of of IN web-archive-org-7310 337 17 a a DT web-archive-org-7310 337 18 web web NN web-archive-org-7310 337 19 page page NN web-archive-org-7310 337 20 collection collection NN web-archive-org-7310 337 21 or or CC web-archive-org-7310 337 22 index index NN web-archive-org-7310 337 23 . . . web-archive-org-7310 338 1 We -PRON- PRP web-archive-org-7310 338 2 have have VBP web-archive-org-7310 338 3 done do VBN web-archive-org-7310 338 4 some some DT web-archive-org-7310 338 5 limited limited JJ web-archive-org-7310 338 6 experiments experiment NNS web-archive-org-7310 338 7 with with IN web-archive-org-7310 338 8 the the DT web-archive-org-7310 338 9 sampled sample VBN web-archive-org-7310 338 10 full full JJ web-archive-org-7310 338 11 text text NN web-archive-org-7310 338 12 documents document NNS web-archive-org-7310 338 13 to to TO web-archive-org-7310 338 14 investigate investigate VB web-archive-org-7310 338 15 some some DT web-archive-org-7310 338 16 of of IN web-archive-org-7310 338 17 our -PRON- PRP$ web-archive-org-7310 338 18 more more RBR web-archive-org-7310 338 19 perplexing perplexing JJ web-archive-org-7310 338 20 results result NNS web-archive-org-7310 338 21 . . . web-archive-org-7310 339 1 These these DT web-archive-org-7310 339 2 experiments experiment NNS web-archive-org-7310 339 3 helped help VBD web-archive-org-7310 339 4 us -PRON- PRP web-archive-org-7310 339 5 in in IN web-archive-org-7310 339 6 uncovering uncover VBG web-archive-org-7310 339 7 a a DT web-archive-org-7310 339 8 source source NN web-archive-org-7310 339 9 of of IN web-archive-org-7310 339 10 pollution pollution NN web-archive-org-7310 339 11 in in IN web-archive-org-7310 339 12 our -PRON- PRP$ web-archive-org-7310 339 13 data data NN web-archive-org-7310 339 14 set set NN web-archive-org-7310 339 15 , , , web-archive-org-7310 339 16 namely namely RB web-archive-org-7310 339 17 machine machine NN web-archive-org-7310 339 18 - - HYPH web-archive-org-7310 339 19 generated generate VBN web-archive-org-7310 339 20 pages page NNS web-archive-org-7310 339 21 constructed construct VBN web-archive-org-7310 339 22 for for IN web-archive-org-7310 339 23 the the DT web-archive-org-7310 339 24 purpose purpose NN web-archive-org-7310 339 25 of of IN web-archive-org-7310 339 26 spamming spamme VBG web-archive-org-7310 339 27 search search NN web-archive-org-7310 339 28 engines engine NNS web-archive-org-7310 339 29 . . . web-archive-org-7310 340 1 We -PRON- PRP web-archive-org-7310 340 2 hope hope VBP web-archive-org-7310 340 3 that that IN web-archive-org-7310 340 4 future future JJ web-archive-org-7310 340 5 work work NN web-archive-org-7310 340 6 using use VBG web-archive-org-7310 340 7 the the DT web-archive-org-7310 340 8 sampled sample VBN web-archive-org-7310 340 9 full full JJ web-archive-org-7310 340 10 text text NN web-archive-org-7310 340 11 documents document NNS web-archive-org-7310 340 12 will will MD web-archive-org-7310 340 13 provide provide VB web-archive-org-7310 340 14 us -PRON- PRP web-archive-org-7310 340 15 with with IN web-archive-org-7310 340 16 additional additional JJ web-archive-org-7310 340 17 insights insight NNS web-archive-org-7310 340 18 . . . web-archive-org-7310 341 1 9 9 CD web-archive-org-7310 341 2       _SP web-archive-org-7310 341 3 REFERENCES reference NNS web-archive-org-7310 341 4 [ [ -LRB- web-archive-org-7310 341 5 1 1 CD web-archive-org-7310 341 6 ] ] -RRB- web-archive-org-7310 341 7 B. B. NNP web-archive-org-7310 341 8 Brewington Brewington NNP web-archive-org-7310 341 9 , , , web-archive-org-7310 341 10 G. G. NNP web-archive-org-7310 341 11 Cybenko Cybenko NNP web-archive-org-7310 341 12 . . . web-archive-org-7310 342 1 How how WRB web-archive-org-7310 342 2 dynamic dynamic JJ web-archive-org-7310 342 3 is be VBZ web-archive-org-7310 342 4 the the DT web-archive-org-7310 342 5 web web NN web-archive-org-7310 342 6 ? ? . web-archive-org-7310 343 1 In in IN web-archive-org-7310 343 2 Proc Proc NNP web-archive-org-7310 343 3 . . . web-archive-org-7310 344 1 of of IN web-archive-org-7310 344 2 the the DT web-archive-org-7310 344 3 9th 9th JJ web-archive-org-7310 344 4 International International NNP web-archive-org-7310 344 5 World World NNP web-archive-org-7310 344 6 Wide Wide NNP web-archive-org-7310 344 7 Web Web NNP web-archive-org-7310 344 8 Conference Conference NNP web-archive-org-7310 344 9 , , , web-archive-org-7310 344 10 May May NNP web-archive-org-7310 344 11 2000 2000 CD web-archive-org-7310 344 12 . . . web-archive-org-7310 345 1 [ [ -LRB- web-archive-org-7310 345 2 2 2 LS web-archive-org-7310 345 3 ] ] -RRB- web-archive-org-7310 345 4 K. K. NNP web-archive-org-7310 345 5 Bharat Bharat NNP web-archive-org-7310 345 6 and and CC web-archive-org-7310 345 7 A. A. NNP web-archive-org-7310 345 8 Broder Broder NNP web-archive-org-7310 345 9 . . . web-archive-org-7310 346 1 Mirror Mirror NNP web-archive-org-7310 346 2 , , , web-archive-org-7310 346 3 mirror mirror NN web-archive-org-7310 346 4 on on IN web-archive-org-7310 346 5 the the DT web-archive-org-7310 346 6 web web NN web-archive-org-7310 346 7 : : : web-archive-org-7310 346 8 a a DT web-archive-org-7310 346 9 study study NN web-archive-org-7310 346 10 of of IN web-archive-org-7310 346 11 host host NN web-archive-org-7310 346 12 pairs pair NNS web-archive-org-7310 346 13 with with IN web-archive-org-7310 346 14 replicated replicate VBN web-archive-org-7310 346 15 content content NN web-archive-org-7310 346 16 . . . web-archive-org-7310 347 1 In in IN web-archive-org-7310 347 2 Proc Proc NNP web-archive-org-7310 347 3 . . . web-archive-org-7310 348 1 of of IN web-archive-org-7310 348 2 the the DT web-archive-org-7310 348 3 8th 8th JJ web-archive-org-7310 348 4 International International NNP web-archive-org-7310 348 5 World World NNP web-archive-org-7310 348 6 Wide Wide NNP web-archive-org-7310 348 7 Web Web NNP web-archive-org-7310 348 8 Conference Conference NNP web-archive-org-7310 348 9 , , , web-archive-org-7310 348 10 May May NNP web-archive-org-7310 348 11 1999 1999 CD web-archive-org-7310 348 12 . . . web-archive-org-7310 349 1 [ [ -LRB- web-archive-org-7310 349 2 3 3 LS web-archive-org-7310 349 3 ] ] -RRB- web-archive-org-7310 349 4 A. a. NN web-archive-org-7310 349 5 Broder Broder NNP web-archive-org-7310 349 6 . . . web-archive-org-7310 350 1 Some some DT web-archive-org-7310 350 2 applications application NNS web-archive-org-7310 350 3 of of IN web-archive-org-7310 350 4 Rabin Rabin NNP web-archive-org-7310 350 5 's 's POS web-archive-org-7310 350 6 fingerprinting fingerprint VBG web-archive-org-7310 350 7 method method NN web-archive-org-7310 350 8 . . . web-archive-org-7310 351 1 In in IN web-archive-org-7310 351 2 R. R. NNP web-archive-org-7310 351 3 Capocelli Capocelli NNP web-archive-org-7310 351 4 , , , web-archive-org-7310 351 5 A. a. NN web-archive-org-7310 352 1 De De NNP web-archive-org-7310 352 2 Santis Santis NNP web-archive-org-7310 352 3 , , , web-archive-org-7310 352 4 and and CC web-archive-org-7310 352 5 U. U. NNP web-archive-org-7310 352 6 Vaccaro Vaccaro NNP web-archive-org-7310 352 7 , , , web-archive-org-7310 352 8 editors editor NNS web-archive-org-7310 352 9 , , , web-archive-org-7310 352 10 Sequences Sequences NNP web-archive-org-7310 352 11 II ii PRP web-archive-org-7310 352 12 : : : web-archive-org-7310 352 13 Methods method NNS web-archive-org-7310 352 14 in in IN web-archive-org-7310 352 15 Communications Communications NNPS web-archive-org-7310 352 16 , , , web-archive-org-7310 352 17 Security Security NNP web-archive-org-7310 352 18 , , , web-archive-org-7310 352 19 and and CC web-archive-org-7310 352 20 Computer Computer NNP web-archive-org-7310 352 21 Science Science NNP web-archive-org-7310 352 22 , , , web-archive-org-7310 352 23 Springer Springer NNP web-archive-org-7310 352 24 - - HYPH web-archive-org-7310 352 25 Verlag Verlag NNP web-archive-org-7310 352 26 , , , web-archive-org-7310 352 27 1993 1993 CD web-archive-org-7310 352 28 . . . web-archive-org-7310 353 1 [ [ -LRB- web-archive-org-7310 353 2 4 4 CD web-archive-org-7310 353 3 ] ] -RRB- web-archive-org-7310 353 4 A. a. NN web-archive-org-7310 353 5 Broder Broder NNP web-archive-org-7310 353 6 , , , web-archive-org-7310 353 7 S. S. NNP web-archive-org-7310 353 8 Glassman Glassman NNP web-archive-org-7310 353 9 , , , web-archive-org-7310 353 10 M. M. NNP web-archive-org-7310 353 11 Manasse Manasse NNP web-archive-org-7310 353 12 , , , web-archive-org-7310 353 13 and and CC web-archive-org-7310 353 14 G. G. NNP web-archive-org-7310 353 15 Zweig Zweig NNP web-archive-org-7310 353 16 . . . web-archive-org-7310 354 1 Syntactic syntactic JJ web-archive-org-7310 354 2 clustering clustering NN web-archive-org-7310 354 3 of of IN web-archive-org-7310 354 4 the the DT web-archive-org-7310 354 5 web web NN web-archive-org-7310 354 6 . . . web-archive-org-7310 355 1 In in IN web-archive-org-7310 355 2 Proc Proc NNP web-archive-org-7310 355 3 . . . web-archive-org-7310 356 1 of of IN web-archive-org-7310 356 2 the the DT web-archive-org-7310 356 3 6th 6th JJ web-archive-org-7310 356 4 International International NNP web-archive-org-7310 356 5 World World NNP web-archive-org-7310 356 6 Wide Wide NNP web-archive-org-7310 356 7 Web Web NNP web-archive-org-7310 356 8 Conference Conference NNP web-archive-org-7310 356 9 , , , web-archive-org-7310 356 10 Apr. April NNP web-archive-org-7310 357 1 1997 1997 CD web-archive-org-7310 357 2 . . . web-archive-org-7310 358 1 [ [ -LRB- web-archive-org-7310 358 2 5 5 CD web-archive-org-7310 358 3 ] ] -RRB- web-archive-org-7310 358 4 J. J. NNP web-archive-org-7310 358 5 Cho Cho NNP web-archive-org-7310 358 6 and and CC web-archive-org-7310 358 7 H. H. NNP web-archive-org-7310 358 8 Garcia Garcia NNP web-archive-org-7310 358 9 - - HYPH web-archive-org-7310 358 10 Molina Molina NNP web-archive-org-7310 358 11 . . . web-archive-org-7310 359 1 The the DT web-archive-org-7310 359 2 evolution evolution NN web-archive-org-7310 359 3 of of IN web-archive-org-7310 359 4 the the DT web-archive-org-7310 359 5 web web NN web-archive-org-7310 359 6 and and CC web-archive-org-7310 359 7 implications implication NNS web-archive-org-7310 359 8 for for IN web-archive-org-7310 359 9 an an DT web-archive-org-7310 359 10 incremental incremental JJ web-archive-org-7310 359 11 crawler crawler NN web-archive-org-7310 359 12 . . . web-archive-org-7310 360 1 In in IN web-archive-org-7310 360 2 Proc Proc NNP web-archive-org-7310 360 3 . . . web-archive-org-7310 361 1 of of IN web-archive-org-7310 361 2 the the DT web-archive-org-7310 361 3 26th 26th JJ web-archive-org-7310 361 4 International International NNP web-archive-org-7310 361 5 Conference Conference NNP web-archive-org-7310 361 6 on on IN web-archive-org-7310 361 7 Very very RB web-archive-org-7310 361 8 Large large JJ web-archive-org-7310 361 9 Databases database NNS web-archive-org-7310 361 10 , , , web-archive-org-7310 361 11 Sep. September NNP web-archive-org-7310 361 12 2000 2000 CD web-archive-org-7310 361 13 . . . web-archive-org-7310 362 1 [ [ -LRB- web-archive-org-7310 362 2 6 6 CD web-archive-org-7310 362 3 ] ] -RRB- web-archive-org-7310 362 4 F. F. NNP web-archive-org-7310 362 5 Douglis Douglis NNP web-archive-org-7310 362 6 , , , web-archive-org-7310 362 7 A. a. NN web-archive-org-7310 362 8 Feldmann Feldmann NNP web-archive-org-7310 362 9 , , , web-archive-org-7310 362 10 B. B. NNP web-archive-org-7310 362 11 Krishnamurthy Krishnamurthy NNP web-archive-org-7310 362 12 , , , web-archive-org-7310 362 13 J. J. NNP web-archive-org-7310 363 1 Mogul Mogul NNP web-archive-org-7310 363 2 . . . web-archive-org-7310 364 1 Rate rate NN web-archive-org-7310 364 2 of of IN web-archive-org-7310 364 3 change change NN web-archive-org-7310 364 4 and and CC web-archive-org-7310 364 5 other other JJ web-archive-org-7310 364 6 metrics metric NNS web-archive-org-7310 364 7 : : : web-archive-org-7310 364 8 a a DT web-archive-org-7310 364 9 live live JJ web-archive-org-7310 364 10 study study NN web-archive-org-7310 364 11 of of IN web-archive-org-7310 364 12 the the DT web-archive-org-7310 364 13 world world NN web-archive-org-7310 364 14 wide wide JJ web-archive-org-7310 364 15 web web NN web-archive-org-7310 364 16 . . . web-archive-org-7310 365 1 In in IN web-archive-org-7310 365 2 USENIX USENIX NNP web-archive-org-7310 365 3 Symposium Symposium NNP web-archive-org-7310 365 4 on on IN web-archive-org-7310 365 5 Internetworking Internetworking NNP web-archive-org-7310 365 6 Technologies Technologies NNPS web-archive-org-7310 365 7 and and CC web-archive-org-7310 365 8 Systems Systems NNPS web-archive-org-7310 365 9 , , , web-archive-org-7310 365 10 Dec. December NNP web-archive-org-7310 365 11 1997 1997 CD web-archive-org-7310 365 12 . . . web-archive-org-7310 366 1 [ [ -LRB- web-archive-org-7310 366 2 7 7 CD web-archive-org-7310 366 3 ] ] -RRB- web-archive-org-7310 366 4 Google Google NNP web-archive-org-7310 366 5 Information Information NNP web-archive-org-7310 366 6 for for IN web-archive-org-7310 366 7 Webmasters webmaster NNS web-archive-org-7310 366 8 . . . web-archive-org-7310 367 1 http://www.google.com/webmasters/2.html http://www.google.com/webmasters/2.html NNP web-archive-org-7310 367 2 [ [ -LRB- web-archive-org-7310 367 3 8 8 CD web-archive-org-7310 367 4 ] ] -RRB- web-archive-org-7310 367 5 M. M. NNP web-archive-org-7310 367 6 Najork Najork NNP web-archive-org-7310 367 7 and and CC web-archive-org-7310 367 8 A. A. NNP web-archive-org-7310 367 9 Heydon Heydon NNP web-archive-org-7310 367 10 . . . web-archive-org-7310 368 1 High high JJ web-archive-org-7310 368 2 - - HYPH web-archive-org-7310 368 3 performance performance NN web-archive-org-7310 368 4 web web NN web-archive-org-7310 368 5 crawling crawling NN web-archive-org-7310 368 6 . . . web-archive-org-7310 369 1 SRC SRC NNP web-archive-org-7310 369 2 Research Research NNP web-archive-org-7310 369 3 Report Report NNP web-archive-org-7310 369 4 173 173 CD web-archive-org-7310 369 5 , , , web-archive-org-7310 369 6 Compaq Compaq NNP web-archive-org-7310 369 7 Systems Systems NNPS web-archive-org-7310 369 8 Research Research NNP web-archive-org-7310 369 9 Center Center NNP web-archive-org-7310 369 10 , , , web-archive-org-7310 369 11 Palo Palo NNP web-archive-org-7310 369 12 Alto Alto NNP web-archive-org-7310 369 13 , , , web-archive-org-7310 369 14 CA CA NNP web-archive-org-7310 369 15 , , , web-archive-org-7310 369 16 Sep. September NNP web-archive-org-7310 369 17 2001 2001 CD web-archive-org-7310 369 18 . . . web-archive-org-7310 370 1 [ [ -LRB- web-archive-org-7310 370 2 9 9 CD web-archive-org-7310 370 3 ] ] -RRB- web-archive-org-7310 370 4 M. M. NNP web-archive-org-7310 370 5 Najork Najork NNP web-archive-org-7310 370 6 and and CC web-archive-org-7310 370 7 J. J. NNP web-archive-org-7310 370 8 Wiener Wiener NNP web-archive-org-7310 370 9 . . . web-archive-org-7310 371 1 Breadth breadth JJ web-archive-org-7310 371 2 - - HYPH web-archive-org-7310 371 3 first first JJ web-archive-org-7310 371 4 search search NN web-archive-org-7310 371 5 crawling crawling NN web-archive-org-7310 371 6 yields yield NNS web-archive-org-7310 371 7 high high JJ web-archive-org-7310 371 8 - - HYPH web-archive-org-7310 371 9 quality quality NN web-archive-org-7310 371 10 pages page NNS web-archive-org-7310 371 11 . . . web-archive-org-7310 372 1 In in IN web-archive-org-7310 372 2 Proc Proc NNP web-archive-org-7310 372 3 . . . web-archive-org-7310 373 1 of of IN web-archive-org-7310 373 2 the the DT web-archive-org-7310 373 3 10th 10th JJ web-archive-org-7310 373 4 International International NNP web-archive-org-7310 373 5 World World NNP web-archive-org-7310 373 6 Wide Wide NNP web-archive-org-7310 373 7 Web Web NNP web-archive-org-7310 373 8 Conference Conference NNP web-archive-org-7310 373 9 , , , web-archive-org-7310 373 10 May May NNP web-archive-org-7310 373 11 2001 2001 CD web-archive-org-7310 373 12 . . . web-archive-org-7310 374 1 [ [ -LRB- web-archive-org-7310 374 2 10 10 CD web-archive-org-7310 374 3 ] ] -RRB- web-archive-org-7310 374 4 M. M. NNP web-archive-org-7310 374 5 Rabin Rabin NNP web-archive-org-7310 374 6 . . . web-archive-org-7310 375 1 Fingerprinting fingerprint VBG web-archive-org-7310 375 2 by by IN web-archive-org-7310 375 3 random random JJ web-archive-org-7310 375 4 polynomials polynomial NNS web-archive-org-7310 375 5 . . . web-archive-org-7310 376 1 Report Report NNP web-archive-org-7310 376 2 TR-15 TR-15 NNP web-archive-org-7310 376 3 - - HYPH web-archive-org-7310 376 4 81 81 CD web-archive-org-7310 376 5 , , , web-archive-org-7310 376 6 Center Center NNP web-archive-org-7310 376 7 for for IN web-archive-org-7310 376 8 Research Research NNP web-archive-org-7310 376 9 in in IN web-archive-org-7310 376 10 Computing Computing NNP web-archive-org-7310 376 11 Technology Technology NNP web-archive-org-7310 376 12 , , , web-archive-org-7310 376 13 Harvard Harvard NNP web-archive-org-7310 376 14 University University NNP web-archive-org-7310 376 15 , , , web-archive-org-7310 376 16 1981 1981 CD web-archive-org-7310 376 17 . . . web-archive-org-7310 377 1 [ [ -LRB- web-archive-org-7310 377 2 11 11 CD web-archive-org-7310 377 3 ] ] -RRB- web-archive-org-7310 377 4 Q. q. NN web-archive-org-7310 378 1 Sun Sun NNP web-archive-org-7310 378 2 , , , web-archive-org-7310 378 3 D. D. NNP web-archive-org-7310 378 4 Simon Simon NNP web-archive-org-7310 378 5 , , , web-archive-org-7310 378 6 Y. Y. NNP web-archive-org-7310 378 7 Wang Wang NNP web-archive-org-7310 378 8 , , , web-archive-org-7310 378 9 W. W. NNP web-archive-org-7310 378 10 Russell Russell NNP web-archive-org-7310 378 11 , , , web-archive-org-7310 378 12 V. V. NNP web-archive-org-7310 378 13 Padmanabhan Padmanabhan NNP web-archive-org-7310 378 14 , , , web-archive-org-7310 378 15 and and CC web-archive-org-7310 378 16 L. L. NNP web-archive-org-7310 378 17 Qiu Qiu NNP web-archive-org-7310 378 18 . . . web-archive-org-7310 379 1 Statistical statistical JJ web-archive-org-7310 379 2 identification identification NN web-archive-org-7310 379 3 of of IN web-archive-org-7310 379 4 encrypted encrypt VBN web-archive-org-7310 379 5 web web NN web-archive-org-7310 379 6 browsing browse VBG web-archive-org-7310 379 7 traffic traffic NN web-archive-org-7310 379 8 . . . web-archive-org-7310 380 1 In in IN web-archive-org-7310 380 2 Proc Proc NNP web-archive-org-7310 380 3 . . . web-archive-org-7310 381 1 of of IN web-archive-org-7310 381 2 the the DT web-archive-org-7310 381 3 IEEE IEEE NNP web-archive-org-7310 381 4 Symposium Symposium NNP web-archive-org-7310 381 5 on on IN web-archive-org-7310 381 6 Security Security NNP web-archive-org-7310 381 7 and and CC web-archive-org-7310 381 8 Privacy Privacy NNP web-archive-org-7310 381 9 , , , web-archive-org-7310 381 10 May May NNP web-archive-org-7310 381 11 2002 2002 CD web-archive-org-7310 381 12 . . . web-archive-org-7310 382 1 Footnotes footnote NNS web-archive-org-7310 382 2 : : : web-archive-org-7310 382 3 1 1 LS web-archive-org-7310 382 4   _SP web-archive-org-7310 382 5 This this DT web-archive-org-7310 382 6 definition definition NN web-archive-org-7310 382 7 of of IN web-archive-org-7310 382 8 identity identity NN web-archive-org-7310 382 9 differs differ NNS web-archive-org-7310 382 10 from from IN web-archive-org-7310 382 11 the the DT web-archive-org-7310 382 12 standard standard JJ web-archive-org-7310 382 13 definition definition NN web-archive-org-7310 382 14 . . . web-archive-org-7310 383 1 For for IN web-archive-org-7310 383 2 example example NN web-archive-org-7310 383 3 , , , web-archive-org-7310 383 4 a a DT web-archive-org-7310 383 5 document document NN web-archive-org-7310 383 6 A A NNP web-archive-org-7310 383 7 is be VBZ web-archive-org-7310 383 8 considered consider VBN web-archive-org-7310 383 9 identical identical JJ web-archive-org-7310 383 10 to to IN web-archive-org-7310 383 11 AA AA NNP web-archive-org-7310 383 12 , , , web-archive-org-7310 383 13 the the DT web-archive-org-7310 383 14 concatenation concatenation NN web-archive-org-7310 383 15 of of IN web-archive-org-7310 383 16 A A NNP web-archive-org-7310 383 17 to to IN web-archive-org-7310 383 18 itself -PRON- PRP web-archive-org-7310 383 19 . . . web-archive-org-7310 384 1 Similarly similarly RB web-archive-org-7310 384 2 , , , web-archive-org-7310 384 3 given give VBN web-archive-org-7310 384 4 two two CD web-archive-org-7310 384 5 documents document NNS web-archive-org-7310 384 6 A a NN web-archive-org-7310 384 7 and and CC web-archive-org-7310 384 8 B b NN web-archive-org-7310 384 9 , , , web-archive-org-7310 384 10 document document NN web-archive-org-7310 384 11 AB AB NNP web-archive-org-7310 384 12 is be VBZ web-archive-org-7310 384 13 identical identical JJ web-archive-org-7310 384 14 to to TO web-archive-org-7310 384 15 document document VB web-archive-org-7310 384 16 BA BA NNP web-archive-org-7310 384 17 . . . web-archive-org-7310 385 1 2 2 LS web-archive-org-7310 385 2   _SP web-archive-org-7310 385 3 We -PRON- PRP web-archive-org-7310 385 4 drew draw VBD web-archive-org-7310 385 5 the the DT web-archive-org-7310 385 6 84 84 CD web-archive-org-7310 385 7 functions function NNS web-archive-org-7310 385 8 from from IN web-archive-org-7310 385 9 a a DT web-archive-org-7310 385 10 smaller small JJR web-archive-org-7310 385 11 space space NN web-archive-org-7310 385 12 , , , web-archive-org-7310 385 13 which which WDT web-archive-org-7310 385 14 does do VBZ web-archive-org-7310 385 15 not not RB web-archive-org-7310 385 16 seem seem VB web-archive-org-7310 385 17 to to TO web-archive-org-7310 385 18 be be VB web-archive-org-7310 385 19 a a DT web-archive-org-7310 385 20 problem problem NN web-archive-org-7310 385 21 in in IN web-archive-org-7310 385 22 practice practice NN web-archive-org-7310 385 23 . . . web-archive-org-7310 386 1 Also also RB web-archive-org-7310 386 2 , , , web-archive-org-7310 386 3 because because IN web-archive-org-7310 386 4 we -PRON- PRP web-archive-org-7310 386 5 fixed fix VBD web-archive-org-7310 386 6 them -PRON- PRP web-archive-org-7310 386 7 at at IN web-archive-org-7310 386 8 the the DT web-archive-org-7310 386 9 time time NN web-archive-org-7310 386 10 when when WRB web-archive-org-7310 386 11 we -PRON- PRP web-archive-org-7310 386 12 wrote write VBD web-archive-org-7310 386 13 the the DT web-archive-org-7310 386 14 feature feature NN web-archive-org-7310 386 15 vector vector NN web-archive-org-7310 386 16 extraction extraction NN web-archive-org-7310 386 17 module module NNP web-archive-org-7310 386 18 , , , web-archive-org-7310 386 19 an an DT web-archive-org-7310 386 20 adversary adversary NN web-archive-org-7310 386 21 who who WP web-archive-org-7310 386 22 had have VBD web-archive-org-7310 386 23 the the DT web-archive-org-7310 386 24 code code NN web-archive-org-7310 386 25 of of IN web-archive-org-7310 386 26 that that DT web-archive-org-7310 386 27 module module NN web-archive-org-7310 386 28 could could MD web-archive-org-7310 386 29 produce produce VB web-archive-org-7310 386 30 documents document NNS web-archive-org-7310 386 31 that that WDT web-archive-org-7310 386 32 would would MD web-archive-org-7310 386 33 fool fool VB web-archive-org-7310 386 34 our -PRON- PRP$ web-archive-org-7310 386 35 metric metric NN web-archive-org-7310 386 36 . . . web-archive-org-7310 387 1 3 3 LS web-archive-org-7310 387 2   _SP web-archive-org-7310 387 3 We -PRON- PRP web-archive-org-7310 387 4 actually actually RB web-archive-org-7310 387 5 did do VBD web-archive-org-7310 387 6 encounter encounter VB web-archive-org-7310 387 7 several several JJ web-archive-org-7310 387 8 documents document NNS web-archive-org-7310 387 9 whose whose WP$ web-archive-org-7310 387 10 content content NN web-archive-org-7310 387 11 type type NN web-archive-org-7310 387 12 changed change VBN web-archive-org-7310 387 13 during during IN web-archive-org-7310 387 14 the the DT web-archive-org-7310 387 15 time time NN web-archive-org-7310 387 16 we -PRON- PRP web-archive-org-7310 387 17 observed observe VBD web-archive-org-7310 387 18 them -PRON- PRP web-archive-org-7310 387 19 . . .