id sid tid token lemma pos gm80ht2731s 1 1 data data NOUN gm80ht2731s 1 2 stream stream NOUN gm80ht2731s 1 3 summaries summary NOUN gm80ht2731s 1 4 capture capture VERB gm80ht2731s 1 5 the the DET gm80ht2731s 1 6 essential essential ADJ gm80ht2731s 1 7 characteristics characteristic NOUN gm80ht2731s 1 8 of of ADP gm80ht2731s 1 9 thestream thestream NOUN gm80ht2731s 1 10 in in ADP gm80ht2731s 1 11 a a DET gm80ht2731s 1 12 managable managable ADJ gm80ht2731s 1 13 amount amount NOUN gm80ht2731s 1 14 of of ADP gm80ht2731s 1 15 space.in space.in NUM gm80ht2731s 1 16 this this DET gm80ht2731s 1 17 thesis thesis NOUN gm80ht2731s 1 18 , , PUNCT gm80ht2731s 1 19 we we PRON gm80ht2731s 1 20 evaluatethree evaluatethree VERB gm80ht2731s 1 21 data datum NOUN gm80ht2731s 1 22 stream stream NOUN gm80ht2731s 1 23 summarization summarization NOUN gm80ht2731s 1 24 methods method NOUN gm80ht2731s 1 25 for for ADP gm80ht2731s 1 26 applications application NOUN gm80ht2731s 1 27 in in ADP gm80ht2731s 1 28 machinelearning machinelearning NOUN gm80ht2731s 1 29 . . PUNCT gm80ht2731s 2 1 one one NUM gm80ht2731s 2 2 summary summary NOUN gm80ht2731s 2 3 is be AUX gm80ht2731s 2 4 based base VERB gm80ht2731s 2 5 on on ADP gm80ht2731s 2 6 equal equal ADJ gm80ht2731s 2 7 width width NOUN gm80ht2731s 2 8 binning binning NOUN gm80ht2731s 2 9 and and CCONJ gm80ht2731s 2 10 is be AUX gm80ht2731s 2 11 verysimple verysimple NOUN gm80ht2731s 2 12 to to PART gm80ht2731s 2 13 implement implement VERB gm80ht2731s 2 14 . . PUNCT gm80ht2731s 3 1 the the DET gm80ht2731s 3 2 other other ADJ gm80ht2731s 3 3 two two NUM gm80ht2731s 3 4 are be AUX gm80ht2731s 3 5 based base VERB gm80ht2731s 3 6 on on ADP gm80ht2731s 3 7 approximate approximate ADJ gm80ht2731s 3 8 quantiles quantile NOUN gm80ht2731s 3 9 , , PUNCT gm80ht2731s 3 10 one one NUM gm80ht2731s 3 11 of of ADP gm80ht2731s 3 12 which which PRON gm80ht2731s 3 13 allows allow VERB gm80ht2731s 3 14 us we PRON gm80ht2731s 3 15 to to PART gm80ht2731s 3 16 make make VERB gm80ht2731s 3 17 information information NOUN gm80ht2731s 3 18 entropy entropy NOUN gm80ht2731s 3 19 computations computation NOUN gm80ht2731s 3 20 within within ADP gm80ht2731s 3 21 specified specified ADJ gm80ht2731s 3 22 errorbounds errorbound NOUN gm80ht2731s 3 23 with with ADP gm80ht2731s 3 24 high high ADJ gm80ht2731s 3 25 probability probability NOUN gm80ht2731s 3 26 . . PUNCT gm80ht2731s 4 1 using use VERB gm80ht2731s 4 2 popular popular ADJ gm80ht2731s 4 3 techniques technique NOUN gm80ht2731s 4 4 frommachine frommachine NOUN gm80ht2731s 4 5 learning learn VERB gm80ht2731s 4 6 --- --- PUNCT gm80ht2731s 4 7 information information NOUN gm80ht2731s 4 8 gain gain NOUN gm80ht2731s 4 9 , , PUNCT gm80ht2731s 4 10 gain gain VERB gm80ht2731s 4 11 ratio ratio NOUN gm80ht2731s 4 12 , , PUNCT gm80ht2731s 4 13 andnaive andnaive ADJ gm80ht2731s 4 14 - - PUNCT gm80ht2731s 4 15 bayes bayes NOUN gm80ht2731s 4 16 classification classification NOUN gm80ht2731s 4 17 --- --- PUNCT gm80ht2731s 4 18 we we PRON gm80ht2731s 4 19 evaluate evaluate VERB gm80ht2731s 4 20 thesesummaries thesesummarie NOUN gm80ht2731s 4 21 for for ADP gm80ht2731s 4 22 accuracy accuracy NOUN gm80ht2731s 4 23 and and CCONJ gm80ht2731s 4 24 memory memory NOUN gm80ht2731s 4 25 utilization utilization NOUN gm80ht2731s 4 26 against against ADP gm80ht2731s 4 27 offlinecomputations offlinecomputation NOUN gm80ht2731s 4 28 that that PRON gm80ht2731s 4 29 assume assume VERB gm80ht2731s 4 30 all all DET gm80ht2731s 4 31 data datum NOUN gm80ht2731s 4 32 is be AUX gm80ht2731s 4 33 available available ADJ gm80ht2731s 4 34 for for ADP gm80ht2731s 4 35 repeatedaccess repeatedaccess NOUN gm80ht2731s 4 36 . . PUNCT gm80ht2731s 5 1 our our PRON gm80ht2731s 5 2 results result NOUN gm80ht2731s 5 3 indicate indicate VERB gm80ht2731s 5 4 that that SCONJ gm80ht2731s 5 5 while while SCONJ gm80ht2731s 5 6 summarization summarization NOUN gm80ht2731s 5 7 based base VERB gm80ht2731s 5 8 on on ADP gm80ht2731s 5 9 equalwidth equalwidth ADJ gm80ht2731s 5 10 binning binning NOUN gm80ht2731s 5 11 performs perform VERB gm80ht2731s 5 12 very very ADV gm80ht2731s 5 13 well well ADV gm80ht2731s 5 14 on on ADP gm80ht2731s 5 15 data datum NOUN gm80ht2731s 5 16 streams stream VERB gm80ht2731s 5 17 with with ADP gm80ht2731s 5 18 a a DET gm80ht2731s 5 19 stationarydistribution stationarydistribution NOUN gm80ht2731s 5 20 , , PUNCT gm80ht2731s 5 21 performance performance NOUN gm80ht2731s 5 22 degrades degrade VERB gm80ht2731s 5 23 on on ADP gm80ht2731s 5 24 certain certain ADJ gm80ht2731s 5 25 non non ADJ gm80ht2731s 5 26 - - ADJ gm80ht2731s 5 27 stationary stationary ADJ gm80ht2731s 5 28 distributions distribution NOUN gm80ht2731s 5 29 . . PUNCT gm80ht2731s 6 1 in in ADP gm80ht2731s 6 2 constrast constrast PROPN gm80ht2731s 6 3 , , PUNCT gm80ht2731s 6 4 summarizations summarization NOUN gm80ht2731s 6 5 based base VERB gm80ht2731s 6 6 on on ADP gm80ht2731s 6 7 approximate approximate ADJ gm80ht2731s 6 8 quantiles quantile NOUN gm80ht2731s 6 9 are be AUX gm80ht2731s 6 10 consistently consistently ADV gm80ht2731s 6 11 close close ADJ gm80ht2731s 6 12 to to ADP gm80ht2731s 6 13 the the DET gm80ht2731s 6 14 offline offline ADJ gm80ht2731s 6 15 values value NOUN gm80ht2731s 6 16 for for ADP gm80ht2731s 6 17 a a DET gm80ht2731s 6 18 variety variety NOUN gm80ht2731s 6 19 of of ADP gm80ht2731s 6 20 stationary stationary ADJ gm80ht2731s 6 21 and and CCONJ gm80ht2731s 6 22 non non ADJ gm80ht2731s 6 23 - - NOUN gm80ht2731s 6 24 stationarydistributions stationarydistribution NOUN gm80ht2731s 6 25 . . PUNCT