id	sid	tid	token	lemma	pos
3262	1	1	aRtiCLE	aRtiCLE	NFP
3262	1	2	titLE	title	CD
3262	1	3	|	|	NNP
3262	1	4	autHoR	autHoR	NNP
3262	1	5	41ContEnt	41ContEnt	NNP
3262	1	6	-	-	HYPH
3262	1	7	BasED	BasED	NNP
3262	1	8	inFoRmation	information	NN
3262	1	9	REtRiEvaL	REtRiEvaL	NNP
3262	1	10	anD	and	CC
3262	1	11	DiGitaL	DiGitaL	NNP
3262	1	12	LiBRaRiEs	LiBRaRiEs	NNP
3262	1	13	|	|	NNP
3262	1	14	Wan	Wan	NNP
3262	1	15	anD	and	CC
3262	1	16	Liu	Liu	NNP
3262	1	17	41	41	CD
3262	1	18	Content	content	NN
3262	1	19	-	-	HYPH
3262	1	20	Based	Based	NNP
3262	1	21	Information	Information	NNP
3262	1	22	Retrieval	Retrieval	NNP
3262	1	23	and	and	CC
3262	1	24	Digital	Digital	NNP
3262	1	25	Libraries	Libraries	NNPS
3262	1	26	This	this	DT
3262	1	27	paper	paper	NN
3262	1	28	discusses	discuss	VBZ
3262	1	29	the	the	DT
3262	1	30	applications	application	NNS
3262	1	31	and	and	CC
3262	1	32	importance	importance	NN
3262	1	33	of	of	IN
3262	1	34	content	content	NN
3262	1	35	-	-	HYPH
3262	1	36	based	base	VBN
3262	1	37	information	information	NN
3262	1	38	retrieval	retrieval	NN
3262	1	39	technology	technology	NN
3262	1	40	in	in	IN
3262	1	41	digital	digital	JJ
3262	1	42	libraries	library	NNS
3262	1	43	.	.	.
3262	2	1	It	-PRON-	PRP
3262	2	2	generalizes	generalize	VBZ
3262	2	3	the	the	DT
3262	2	4	process	process	NN
3262	2	5	and	and	CC
3262	2	6	ana-	ana-	NNP
3262	2	7	lyzes	lyze	VBZ
3262	2	8	current	current	JJ
3262	2	9	examples	example	NNS
3262	2	10	in	in	IN
3262	2	11	four	four	CD
3262	2	12	areas	area	NNS
3262	2	13	of	of	IN
3262	2	14	the	the	DT
3262	2	15	technology	technology	NN
3262	2	16	.	.	.
3262	3	1	Content	content	NN
3262	3	2	-	-	HYPH
3262	3	3	based	base	VBN
3262	3	4	information	information	NN
3262	3	5	retrieval	retrieval	NN
3262	3	6	has	have	VBZ
3262	3	7	been	be	VBN
3262	3	8	shown	show	VBN
3262	3	9	to	to	TO
3262	3	10	be	be	VB
3262	3	11	an	an	DT
3262	3	12	effective	effective	JJ
3262	3	13	way	way	NN
3262	3	14	to	to	TO
3262	3	15	search	search	VB
3262	3	16	for	for	IN
3262	3	17	the	the	DT
3262	3	18	type	type	NN
3262	3	19	of	of	IN
3262	3	20	multime-	multime-	NNP
3262	3	21	dia	dia	NN
3262	3	22	documents	document	NNS
3262	3	23	that	that	WDT
3262	3	24	are	be	VBP
3262	3	25	increasingly	increasingly	RB
3262	3	26	stored	store	VBN
3262	3	27	in	in	IN
3262	3	28	digital	digital	JJ
3262	3	29	libraries	library	NNS
3262	3	30	.	.	.
3262	4	1	As	as	IN
3262	4	2	a	a	DT
3262	4	3	good	good	JJ
3262	4	4	complement	complement	NN
3262	4	5	to	to	IN
3262	4	6	traditional	traditional	JJ
3262	4	7	text-	text-	NN
3262	4	8	based	base	VBN
3262	4	9	information	information	NN
3262	4	10	retrieval	retrieval	NN
3262	4	11	technology	technology	NN
3262	4	12	,	,	,
3262	4	13	content	content	NN
3262	4	14	-	-	HYPH
3262	4	15	based	base	VBN
3262	4	16	information	information	NN
3262	4	17	retrieval	retrieval	NN
3262	4	18	will	will	MD
3262	4	19	be	be	VB
3262	4	20	a	a	DT
3262	4	21	significant	significant	JJ
3262	4	22	trend	trend	NN
3262	4	23	for	for	IN
3262	4	24	the	the	DT
3262	4	25	development	development	NN
3262	4	26	of	of	IN
3262	4	27	digital	digital	JJ
3262	4	28	libraries	library	NNS
3262	4	29	.	.	.
3262	5	1	W	w	NN
3262	5	2	ith	ith	NN
3262	5	3	several	several	JJ
3262	5	4	decades	decade	NNS
3262	5	5	of	of	IN
3262	5	6	their	-PRON-	PRP$
3262	5	7	development	development	NN
3262	5	8	,	,	,
3262	5	9	digital	digital	JJ
3262	5	10	libraries	library	NNS
3262	5	11	are	be	VBP
3262	5	12	no	no	RB
3262	5	13	longer	long	RBR
3262	5	14	a	a	DT
3262	5	15	myth	myth	NN
3262	5	16	.	.	.
3262	6	1	In	in	IN
3262	6	2	fact	fact	NN
3262	6	3	,	,	,
3262	6	4	some	some	DT
3262	6	5	gen-	gen-	NN
3262	6	6	eral	eral	JJ
3262	6	7	digital	digital	JJ
3262	6	8	libraries	library	NNS
3262	6	9	such	such	JJ
3262	6	10	as	as	IN
3262	6	11	the	the	DT
3262	6	12	National	National	NNP
3262	6	13	Science	Science	NNP
3262	6	14	Digital	Digital	NNP
3262	6	15	Library	Library	NNP
3262	6	16	(	(	-LRB-
3262	6	17	NSDL	NSDL	NNP
3262	6	18	)	)	-RRB-
3262	6	19	and	and	CC
3262	6	20	the	the	DT
3262	6	21	Internet	Internet	NNP
3262	6	22	Public	Public	NNP
3262	6	23	Library	Library	NNP
3262	6	24	are	be	VBP
3262	6	25	widely	widely	RB
3262	6	26	known	know	VBN
3262	6	27	and	and	CC
3262	6	28	used	use	VBN
3262	6	29	.	.	.
3262	7	1	The	the	DT
3262	7	2	advance	advance	NN
3262	7	3	of	of	IN
3262	7	4	computer	computer	NN
3262	7	5	technology	technology	NN
3262	7	6	makes	make	VBZ
3262	7	7	it	-PRON-	PRP
3262	7	8	possible	possible	JJ
3262	7	9	to	to	TO
3262	7	10	include	include	VB
3262	7	11	a	a	DT
3262	7	12	colossal	colossal	JJ
3262	7	13	amount	amount	NN
3262	7	14	of	of	IN
3262	7	15	information	information	NN
3262	7	16	in	in	IN
3262	7	17	various	various	JJ
3262	7	18	formats	format	NNS
3262	7	19	in	in	IN
3262	7	20	a	a	DT
3262	7	21	digital	digital	JJ
3262	7	22	library	library	NN
3262	7	23	.	.	.
3262	8	1	In	in	IN
3262	8	2	addition	addition	NN
3262	8	3	to	to	IN
3262	8	4	traditional	traditional	JJ
3262	8	5	text	text	NN
3262	8	6	-	-	HYPH
3262	8	7	based	base	VBN
3262	8	8	documents	document	NNS
3262	8	9	such	such	JJ
3262	8	10	as	as	IN
3262	8	11	books	book	NNS
3262	8	12	and	and	CC
3262	8	13	articles	article	NNS
3262	8	14	,	,	,
3262	8	15	other	other	JJ
3262	8	16	types	type	NNS
3262	8	17	of	of	IN
3262	8	18	materials	material	NNS
3262	8	19	—	—	:
3262	8	20	including	include	VBG
3262	8	21	images	image	NNS
3262	8	22	,	,	,
3262	8	23	audio	audio	NN
3262	8	24	,	,	,
3262	8	25	and	and	CC
3262	8	26	video	video	NN
3262	8	27	—	—	:
3262	8	28	can	can	MD
3262	8	29	also	also	RB
3262	8	30	be	be	VB
3262	8	31	easily	easily	RB
3262	8	32	digitized	digitize	VBN
3262	8	33	and	and	CC
3262	8	34	stored	stored	JJ
3262	8	35	.	.	.
3262	9	1	Therefore	therefore	RB
3262	9	2	,	,	,
3262	9	3	how	how	WRB
3262	9	4	to	to	TO
3262	9	5	retrieve	retrieve	VB
3262	9	6	and	and	CC
3262	9	7	present	present	VB
3262	9	8	this	this	DT
3262	9	9	multimedia	multimedia	NNS
3262	9	10	information	information	NN
3262	9	11	effectively	effectively	RB
3262	9	12	through	through	IN
3262	9	13	the	the	DT
3262	9	14	interface	interface	NN
3262	9	15	of	of	IN
3262	9	16	a	a	DT
3262	9	17	digital	digital	JJ
3262	9	18	library	library	NN
3262	9	19	becomes	become	VBZ
3262	9	20	a	a	DT
3262	9	21	significant	significant	JJ
3262	9	22	research	research	NN
3262	9	23	topic	topic	NN
3262	9	24	.	.	.
3262	10	1	Currently	currently	RB
3262	10	2	,	,	,
3262	10	3	there	there	EX
3262	10	4	are	be	VBP
3262	10	5	three	three	CD
3262	10	6	methods	method	NNS
3262	10	7	of	of	IN
3262	10	8	retrieving	retrieve	VBG
3262	10	9	infor-	infor-	XX
3262	10	10	mation	mation	NN
3262	10	11	in	in	IN
3262	10	12	a	a	DT
3262	10	13	digital	digital	JJ
3262	10	14	library	library	NN
3262	10	15	.	.	.
3262	11	1	The	the	DT
3262	11	2	first	first	JJ
3262	11	3	and	and	CC
3262	11	4	the	the	DT
3262	11	5	easiest	easy	JJS
3262	11	6	way	way	NN
3262	11	7	is	be	VBZ
3262	11	8	free	free	JJ
3262	11	9	browsing	browsing	NN
3262	11	10	.	.	.
3262	12	1	By	by	IN
3262	12	2	this	this	DT
3262	12	3	means	mean	VBZ
3262	12	4	,	,	,
3262	12	5	a	a	DT
3262	12	6	user	user	NN
3262	12	7	browses	browse	NNS
3262	12	8	through	through	IN
3262	12	9	a	a	DT
3262	12	10	collection	collection	NN
3262	12	11	and	and	CC
3262	12	12	looks	look	VBZ
3262	12	13	for	for	IN
3262	12	14	desired	desire	VBN
3262	12	15	information	information	NN
3262	12	16	.	.	.
3262	13	1	The	the	DT
3262	13	2	second	second	JJ
3262	13	3	method	method	NN
3262	13	4	—	—	:
3262	13	5	the	the	DT
3262	13	6	most	most	RBS
3262	13	7	popular	popular	JJ
3262	13	8	technique	technique	NN
3262	13	9	used	use	VBN
3262	13	10	today	today	NN
3262	13	11	—	—	:
3262	13	12	is	be	VBZ
3262	13	13	text-	text-	XX
3262	13	14	based	base	VBN
3262	13	15	retrieval	retrieval	NN
3262	13	16	.	.	.
3262	14	1	Through	through	IN
3262	14	2	this	this	DT
3262	14	3	method	method	NN
3262	14	4	,	,	,
3262	14	5	textual	textual	JJ
3262	14	6	information	information	NN
3262	14	7	(	(	-LRB-
3262	14	8	full	full	JJ
3262	14	9	text	text	NN
3262	14	10	of	of	IN
3262	14	11	text	text	NN
3262	14	12	-	-	HYPH
3262	14	13	based	base	VBN
3262	14	14	documents	document	NNS
3262	14	15	and/or	and/or	CC
3262	14	16	metadata	metadata	NN
3262	14	17	of	of	IN
3262	14	18	multimedia	multimedia	NNS
3262	14	19	documents	document	NNS
3262	14	20	)	)	-RRB-
3262	14	21	is	be	VBZ
3262	14	22	indexed	index	VBN
3262	14	23	so	so	IN
3262	14	24	that	that	IN
3262	14	25	a	a	DT
3262	14	26	user	user	NN
3262	14	27	can	can	MD
3262	14	28	search	search	VB
3262	14	29	the	the	DT
3262	14	30	digital	digital	JJ
3262	14	31	library	library	NN
3262	14	32	by	by	IN
3262	14	33	using	use	VBG
3262	14	34	keywords	keyword	NNS
3262	14	35	or	or	CC
3262	14	36	controlled	control	VBN
3262	14	37	terms	term	NNS
3262	14	38	.	.	.
3262	15	1	The	the	DT
3262	15	2	third	third	JJ
3262	15	3	method	method	NN
3262	15	4	is	be	VBZ
3262	15	5	content	content	NN
3262	15	6	-	-	HYPH
3262	15	7	based	base	VBN
3262	15	8	retrieval	retrieval	NN
3262	15	9	,	,	,
3262	15	10	which	which	WDT
3262	15	11	enables	enable	VBZ
3262	15	12	a	a	DT
3262	15	13	user	user	NN
3262	15	14	to	to	TO
3262	15	15	search	search	VB
3262	15	16	multimedia	multimedia	NNS
3262	15	17	information	information	NN
3262	15	18	in	in	IN
3262	15	19	terms	term	NNS
3262	15	20	of	of	IN
3262	15	21	the	the	DT
3262	15	22	actual	actual	JJ
3262	15	23	content	content	NN
3262	15	24	of	of	IN
3262	15	25	image	image	NN
3262	15	26	,	,	,
3262	15	27	audio	audio	NN
3262	15	28	,	,	,
3262	15	29	or	or	CC
3262	15	30	video	video	NN
3262	15	31	(	(	-LRB-
3262	15	32	Marques	Marques	NNP
3262	15	33	and	and	CC
3262	15	34	Furht	Furht	NNP
3262	15	35	2002	2002	CD
3262	15	36	)	)	-RRB-
3262	15	37	.	.	.
3262	16	1	Some	some	DT
3262	16	2	content	content	NN
3262	16	3	features	feature	NNS
3262	16	4	that	that	WDT
3262	16	5	have	have	VBP
3262	16	6	been	be	VBN
3262	16	7	studied	study	VBN
3262	16	8	so	so	RB
3262	16	9	far	far	RB
3262	16	10	include	include	VBP
3262	16	11	color	color	NN
3262	16	12	,	,	,
3262	16	13	texture	texture	NN
3262	16	14	,	,	,
3262	16	15	size	size	NN
3262	16	16	,	,	,
3262	16	17	shape	shape	NN
3262	16	18	,	,	,
3262	16	19	motion	motion	NN
3262	16	20	,	,	,
3262	16	21	and	and	CC
3262	16	22	pitch	pitch	NN
3262	16	23	.	.	.
3262	17	1	While	while	IN
3262	17	2	some	some	DT
3262	17	3	may	may	MD
3262	17	4	argue	argue	VB
3262	17	5	that	that	IN
3262	17	6	text	text	NN
3262	17	7	-	-	HYPH
3262	17	8	based	base	VBN
3262	17	9	retrieval	retrieval	NN
3262	17	10	tech-	tech-	NN
3262	17	11	niques	nique	NNS
3262	17	12	are	be	VBP
3262	17	13	good	good	JJ
3262	17	14	enough	enough	RB
3262	17	15	to	to	TO
3262	17	16	locate	locate	VB
3262	17	17	desired	desire	VBN
3262	17	18	multimedia	multimedia	NNS
3262	17	19	information	information	NN
3262	17	20	,	,	,
3262	17	21	as	as	RB
3262	17	22	long	long	RB
3262	17	23	as	as	IN
3262	17	24	it	-PRON-	PRP
3262	17	25	is	be	VBZ
3262	17	26	assigned	assign	VBN
3262	17	27	proper	proper	JJ
3262	17	28	metadata	metadata	NN
3262	17	29	or	or	CC
3262	17	30	tags	tag	NNS
3262	17	31	,	,	,
3262	17	32	words	word	NNS
3262	17	33	are	be	VBP
3262	17	34	not	not	RB
3262	17	35	sufficient	sufficient	JJ
3262	17	36	to	to	TO
3262	17	37	describe	describe	VB
3262	17	38	what	what	WP
3262	17	39	is	be	VBZ
3262	17	40	some-	some-	NN
3262	17	41	times	time	NNS
3262	17	42	in	in	IN
3262	17	43	a	a	DT
3262	17	44	human	human	NN
3262	17	45	’s	’s	POS
3262	17	46	mind	mind	NN
3262	17	47	.	.	.
3262	18	1	Imagine	imagine	VB
3262	18	2	a	a	DT
3262	18	3	few	few	JJ
3262	18	4	examples	example	NNS
3262	18	5	:	:	:
3262	18	6	A	a	DT
3262	18	7	patron	patron	NN
3262	18	8	comes	come	VBZ
3262	18	9	to	to	IN
3262	18	10	a	a	DT
3262	18	11	public	public	JJ
3262	18	12	library	library	NN
3262	18	13	with	with	IN
3262	18	14	a	a	DT
3262	18	15	picture	picture	NN
3262	18	16	of	of	IN
3262	18	17	a	a	DT
3262	18	18	rare	rare	JJ
3262	18	19	insect	insect	NN
3262	18	20	.	.	.
3262	19	1	Without	without	IN
3262	19	2	expertise	expertise	NN
3262	19	3	in	in	IN
3262	19	4	entomology	entomology	NN
3262	19	5	,	,	,
3262	19	6	the	the	DT
3262	19	7	librarian	librarian	NN
3262	19	8	wo	will	MD
3262	19	9	n’t	not	RB
3262	19	10	know	know	VB
3262	19	11	where	where	WRB
3262	19	12	to	to	TO
3262	19	13	start	start	VB
3262	19	14	if	if	IN
3262	19	15	only	only	RB
3262	19	16	a	a	DT
3262	19	17	text	text	NN
3262	19	18	-	-	HYPH
3262	19	19	based	base	VBN
3262	19	20	informa-	informa-	JJ
3262	19	21	tion	tion	NN
3262	19	22	retrieval	retrieval	NN
3262	19	23	system	system	NN
3262	19	24	is	be	VBZ
3262	19	25	available	available	JJ
3262	19	26	.	.	.
3262	20	1	However	however	RB
3262	20	2	,	,	,
3262	20	3	with	with	IN
3262	20	4	the	the	DT
3262	20	5	help	help	NN
3262	20	6	of	of	IN
3262	20	7	content	content	NN
3262	20	8	-	-	HYPH
3262	20	9	based	base	VBN
3262	20	10	image	image	NN
3262	20	11	retrieval	retrieval	NN
3262	20	12	,	,	,
3262	20	13	the	the	DT
3262	20	14	librarian	librarian	NN
3262	20	15	can	can	MD
3262	20	16	upload	upload	VB
3262	20	17	the	the	DT
3262	20	18	digitized	digitize	VBN
3262	20	19	image	image	NN
3262	20	20	of	of	IN
3262	20	21	the	the	DT
3262	20	22	insect	insect	NN
3262	20	23	to	to	IN
3262	20	24	an	an	DT
3262	20	25	online	online	JJ
3262	20	26	digital	digital	JJ
3262	20	27	image	image	NN
3262	20	28	library	library	NN
3262	20	29	of	of	IN
3262	20	30	insects	insect	NNS
3262	20	31	,	,	,
3262	20	32	and	and	CC
3262	20	33	the	the	DT
3262	20	34	system	system	NN
3262	20	35	will	will	MD
3262	20	36	retrieve	retrieve	VB
3262	20	37	similar	similar	JJ
3262	20	38	images	image	NNS
3262	20	39	with	with	IN
3262	20	40	detailed	detailed	JJ
3262	20	41	description	description	NN
3262	20	42	of	of	IN
3262	20	43	this	this	DT
3262	20	44	insect	insect	NN
3262	20	45	.	.	.
3262	21	1	Similarly	similarly	RB
3262	21	2	,	,	,
3262	21	3	a	a	DT
3262	21	4	patron	patron	NN
3262	21	5	has	have	VBZ
3262	21	6	a	a	DT
3262	21	7	segment	segment	NN
3262	21	8	of	of	IN
3262	21	9	music	music	NN
3262	21	10	audio	audio	NN
3262	21	11	,	,	,
3262	21	12	about	about	IN
3262	21	13	which	which	WDT
3262	21	14	he	-PRON-	PRP
3262	21	15	or	or	CC
3262	21	16	she	-PRON-	PRP
3262	21	17	knows	know	VBZ
3262	21	18	nothing	nothing	NN
3262	21	19	but	but	IN
3262	21	20	wants	want	VBZ
3262	21	21	to	to	TO
3262	21	22	find	find	VB
3262	21	23	out	out	RP
3262	21	24	more	more	JJR
3262	21	25	.	.	.
3262	22	1	By	by	IN
3262	22	2	using	use	VBG
3262	22	3	the	the	DT
3262	22	4	content	content	NN
3262	22	5	-	-	HYPH
3262	22	6	based	base	VBN
3262	22	7	audio	audio	JJ
3262	22	8	retrieval	retrieval	NN
3262	22	9	system	system	NN
3262	22	10	,	,	,
3262	22	11	the	the	DT
3262	22	12	patron	patron	NN
3262	22	13	can	can	MD
3262	22	14	get	get	VB
3262	22	15	similar	similar	JJ
3262	22	16	audio	audio	JJ
3262	22	17	clips	clip	NNS
3262	22	18	with	with	IN
3262	22	19	detailed	detailed	JJ
3262	22	20	information	information	NN
3262	22	21	from	from	IN
3262	22	22	a	a	DT
3262	22	23	digital	digital	JJ
3262	22	24	music	music	NN
3262	22	25	library	library	NN
3262	22	26	,	,	,
3262	22	27	and	and	CC
3262	22	28	then	then	RB
3262	22	29	listen	listen	VB
3262	22	30	to	to	IN
3262	22	31	them	-PRON-	PRP
3262	22	32	to	to	TO
3262	22	33	find	find	VB
3262	22	34	an	an	DT
3262	22	35	exact	exact	JJ
3262	22	36	match	match	NN
3262	22	37	.	.	.
3262	23	1	This	this	DT
3262	23	2	procedure	procedure	NN
3262	23	3	will	will	MD
3262	23	4	be	be	VB
3262	23	5	much	much	RB
3262	23	6	easier	easy	JJR
3262	23	7	than	than	IN
3262	23	8	doing	do	VBG
3262	23	9	a	a	DT
3262	23	10	search	search	NN
3262	23	11	on	on	IN
3262	23	12	a	a	DT
3262	23	13	text	text	NN
3262	23	14	-	-	HYPH
3262	23	15	based	base	VBN
3262	23	16	music	music	NN
3262	23	17	search	search	NN
3262	23	18	system	system	NN
3262	23	19	.	.	.
3262	24	1	It	-PRON-	PRP
3262	24	2	is	be	VBZ
3262	24	3	definitely	definitely	RB
3262	24	4	helpful	helpful	JJ
3262	24	5	if	if	IN
3262	24	6	a	a	DT
3262	24	7	user	user	NN
3262	24	8	can	can	MD
3262	24	9	search	search	VB
3262	24	10	this	this	DT
3262	24	11	non	non	JJ
3262	24	12	-	-	JJ
3262	24	13	textual	textual	JJ
3262	24	14	information	information	NN
3262	24	15	by	by	IN
3262	24	16	styles	style	NNS
3262	24	17	and	and	CC
3262	24	18	features	feature	NNS
3262	24	19	.	.	.
3262	25	1	In	in	IN
3262	25	2	addition	addition	NN
3262	25	3	,	,	,
3262	25	4	the	the	DT
3262	25	5	advance	advance	NN
3262	25	6	of	of	IN
3262	25	7	the	the	DT
3262	25	8	World	World	NNP
3262	25	9	Wide	Wide	NNP
3262	25	10	Web	Web	NNP
3262	25	11	brings	bring	VBZ
3262	25	12	some	some	DT
3262	25	13	new	new	JJ
3262	25	14	challenges	challenge	NNS
3262	25	15	to	to	IN
3262	25	16	traditional	traditional	JJ
3262	25	17	text	text	NN
3262	25	18	-	-	HYPH
3262	25	19	based	base	VBN
3262	25	20	information	information	NN
3262	25	21	retrieval	retrieval	NN
3262	25	22	.	.	.
3262	26	1	While	while	IN
3262	26	2	today	today	NN
3262	26	3	’s	’s	POS
3262	26	4	Web	web	NN
3262	26	5	-	-	HYPH
3262	26	6	based	base	VBN
3262	26	7	digital	digital	JJ
3262	26	8	libraries	library	NNS
3262	26	9	can	can	MD
3262	26	10	be	be	VB
3262	26	11	accessed	access	VBN
3262	26	12	around	around	IN
3262	26	13	the	the	DT
3262	26	14	world	world	NN
3262	26	15	,	,	,
3262	26	16	users	user	NNS
3262	26	17	with	with	IN
3262	26	18	different	different	JJ
3262	26	19	language	language	NN
3262	26	20	and	and	CC
3262	26	21	cultural	cultural	JJ
3262	26	22	backgrounds	background	NNS
3262	26	23	may	may	MD
3262	26	24	not	not	RB
3262	26	25	be	be	VB
3262	26	26	able	able	JJ
3262	26	27	to	to	TO
3262	26	28	do	do	VB
3262	26	29	effective	effective	JJ
3262	26	30	keyword	keyword	NN
3262	26	31	searches	search	NNS
3262	26	32	of	of	IN
3262	26	33	these	these	DT
3262	26	34	librar-	librar-	XX
3262	26	35	ies	ie	NNS
3262	26	36	.	.	.
3262	27	1	Content	content	NN
3262	27	2	-	-	HYPH
3262	27	3	based	base	VBN
3262	27	4	information	information	NN
3262	27	5	retrieval	retrieval	NN
3262	27	6	techniques	technique	NNS
3262	27	7	will	will	MD
3262	27	8	increase	increase	VB
3262	27	9	the	the	DT
3262	27	10	accessibility	accessibility	NN
3262	27	11	of	of	IN
3262	27	12	these	these	DT
3262	27	13	digital	digital	JJ
3262	27	14	libraries	library	NNS
3262	27	15	greatly	greatly	RB
3262	27	16	,	,	,
3262	27	17	and	and	CC
3262	27	18	this	this	DT
3262	27	19	is	be	VBZ
3262	27	20	probably	probably	RB
3262	27	21	a	a	DT
3262	27	22	major	major	JJ
3262	27	23	reason	reason	NN
3262	27	24	it	-PRON-	PRP
3262	27	25	has	have	VBZ
3262	27	26	become	become	VBN
3262	27	27	a	a	DT
3262	27	28	hot	hot	JJ
3262	27	29	research	research	NN
3262	27	30	area	area	NN
3262	27	31	in	in	IN
3262	27	32	the	the	DT
3262	27	33	past	past	JJ
3262	27	34	decade	decade	NN
3262	27	35	.	.	.
3262	28	1	Ideally	ideally	RB
3262	28	2	,	,	,
3262	28	3	a	a	DT
3262	28	4	content	content	NN
3262	28	5	-	-	HYPH
3262	28	6	based	base	VBN
3262	28	7	information	information	NN
3262	28	8	retrieval	retrieval	NN
3262	28	9	system	system	NN
3262	28	10	can	can	MD
3262	28	11	understand	understand	VB
3262	28	12	the	the	DT
3262	28	13	multi-	multi-	NNP
3262	28	14	media	media	NNP
3262	28	15	data	data	NNP
3262	28	16	semantically	semantically	RB
3262	28	17	,	,	,
3262	28	18	such	such	JJ
3262	28	19	as	as	IN
3262	28	20	its	-PRON-	PRP$
3262	28	21	objects	object	NNS
3262	28	22	and	and	CC
3262	28	23	categories	category	NNS
3262	28	24	to	to	TO
3262	28	25	which	which	WDT
3262	28	26	it	-PRON-	PRP
3262	28	27	belongs	belong	VBZ
3262	28	28	.	.	.
3262	29	1	Therefore	therefore	RB
3262	29	2	,	,	,
3262	29	3	a	a	DT
3262	29	4	user	user	NN
3262	29	5	is	be	VBZ
3262	29	6	able	able	JJ
3262	29	7	to	to	TO
3262	29	8	submit	submit	VB
3262	29	9	semantic	semantic	JJ
3262	29	10	queries	query	NNS
3262	29	11	and	and	CC
3262	29	12	retrieve	retrieve	VB
3262	29	13	matched	match	VBN
3262	29	14	results	result	NNS
3262	29	15	.	.	.
3262	30	1	However	however	RB
3262	30	2	,	,	,
3262	30	3	a	a	DT
3262	30	4	great	great	JJ
3262	30	5	difficulty	difficulty	NN
3262	30	6	in	in	IN
3262	30	7	the	the	DT
3262	30	8	current	current	JJ
3262	30	9	computer	computer	NN
3262	30	10	technology	technology	NN
3262	30	11	is	be	VBZ
3262	30	12	to	to	TO
3262	30	13	extract	extract	VB
3262	30	14	high	high	JJ
3262	30	15	-	-	HYPH
3262	30	16	level	level	NN
3262	30	17	or	or	CC
3262	30	18	semantic	semantic	JJ
3262	30	19	features	feature	NNS
3262	30	20	of	of	IN
3262	30	21	multimedia	multimedia	NNS
3262	30	22	information	information	NN
3262	30	23	.	.	.
3262	31	1	Most	Most	JJS
3262	31	2	projects	project	NNS
3262	31	3	still	still	RB
3262	31	4	focus	focus	VBP
3262	31	5	on	on	IN
3262	31	6	lower	low	JJR
3262	31	7	-	-	HYPH
3262	31	8	level	level	NN
3262	31	9	fea-	fea-	NNS
3262	31	10	tures	ture	NNS
3262	31	11	,	,	,
3262	31	12	such	such	JJ
3262	31	13	as	as	IN
3262	31	14	color	color	NN
3262	31	15	,	,	,
3262	31	16	texture	texture	NN
3262	31	17	,	,	,
3262	31	18	and	and	CC
3262	31	19	shape	shape	NN
3262	31	20	.	.	.
3262	32	1	Simply	simply	RB
3262	32	2	put	put	VBN
3262	32	3	,	,	,
3262	32	4	a	a	DT
3262	32	5	typical	typical	JJ
3262	32	6	content	content	NN
3262	32	7	-	-	HYPH
3262	32	8	based	base	VBN
3262	32	9	information	information	NN
3262	32	10	retrieval	retrieval	NN
3262	32	11	system	system	NN
3262	32	12	works	work	VBZ
3262	32	13	in	in	IN
3262	32	14	this	this	DT
3262	32	15	way	way	NN
3262	32	16	:	:	:
3262	32	17	First	first	RB
3262	32	18	,	,	,
3262	32	19	for	for	IN
3262	32	20	each	each	DT
3262	32	21	mul-	mul-	NNP
3262	32	22	timedia	timedia	NNP
3262	32	23	file	file	NN
3262	32	24	in	in	IN
3262	32	25	the	the	DT
3262	32	26	database	database	NN
3262	32	27	,	,	,
3262	32	28	certain	certain	JJ
3262	32	29	feature	feature	NN
3262	32	30	information	information	NN
3262	32	31	(	(	-LRB-
3262	32	32	e.g.	e.g.	RB
3262	32	33	,	,	,
3262	32	34	color	color	NN
3262	32	35	,	,	,
3262	32	36	motion	motion	NN
3262	32	37	,	,	,
3262	32	38	or	or	CC
3262	32	39	pitch	pitch	NN
3262	32	40	)	)	-RRB-
3262	32	41	is	be	VBZ
3262	32	42	extracted	extract	VBN
3262	32	43	,	,	,
3262	32	44	indexed	indexed	JJ
3262	32	45	,	,	,
3262	32	46	and	and	CC
3262	32	47	stored	store	VBD
3262	32	48	.	.	.
3262	33	1	Second	second	JJ
3262	33	2	,	,	,
3262	33	3	when	when	WRB
3262	33	4	a	a	DT
3262	33	5	user	user	NN
3262	33	6	composes	compose	VBZ
3262	33	7	a	a	DT
3262	33	8	query	query	NN
3262	33	9	,	,	,
3262	33	10	the	the	DT
3262	33	11	feature	feature	NN
3262	33	12	information	information	NN
3262	33	13	of	of	IN
3262	33	14	the	the	DT
3262	33	15	query	query	NN
3262	33	16	is	be	VBZ
3262	33	17	calculated	calculate	VBN
3262	33	18	as	as	IN
3262	33	19	vectors	vector	NNS
3262	33	20	.	.	.
3262	34	1	Finally	finally	RB
3262	34	2	,	,	,
3262	34	3	the	the	DT
3262	34	4	system	system	NN
3262	34	5	compares	compare	VBZ
3262	34	6	the	the	DT
3262	34	7	similarity	similarity	NN
3262	34	8	between	between	IN
3262	34	9	the	the	DT
3262	34	10	feature	feature	NN
3262	34	11	vectors	vector	NNS
3262	34	12	of	of	IN
3262	34	13	the	the	DT
3262	34	14	query	query	NNP
3262	34	15	and	and	CC
3262	34	16	multimedia	multimedia	NNP
3262	34	17	data	datum	NNS
3262	34	18	,	,	,
3262	34	19	and	and	CC
3262	34	20	retrieves	retrieve	VBZ
3262	34	21	the	the	DT
3262	34	22	best	good	JJS
3262	34	23	matching	matching	NN
3262	34	24	records	record	NNS
3262	34	25	.	.	.
3262	35	1	If	if	IN
3262	35	2	the	the	DT
3262	35	3	user	user	NN
3262	35	4	is	be	VBZ
3262	35	5	not	not	RB
3262	35	6	satisfied	satisfied	JJ
3262	35	7	with	with	IN
3262	35	8	the	the	DT
3262	35	9	retrieved	retrieved	JJ
3262	35	10	records	record	NNS
3262	35	11	,	,	,
3262	35	12	he	-PRON-	PRP
3262	35	13	or	or	CC
3262	35	14	she	-PRON-	PRP
3262	35	15	can	can	MD
3262	35	16	refine	refine	VB
3262	35	17	the	the	DT
3262	35	18	search	search	NN
3262	35	19	results	result	NNS
3262	35	20	by	by	IN
3262	35	21	selecting	select	VBG
3262	35	22	the	the	DT
3262	35	23	most	most	RBS
3262	35	24	relevant	relevant	JJ
3262	35	25	ones	one	NNS
3262	35	26	to	to	IN
3262	35	27	the	the	DT
3262	35	28	search	search	NN
3262	35	29	query	query	NN
3262	35	30	,	,	,
3262	35	31	and	and	CC
3262	35	32	repeat	repeat	VB
3262	35	33	the	the	DT
3262	35	34	search	search	NN
3262	35	35	with	with	IN
3262	35	36	the	the	DT
3262	35	37	new	new	JJ
3262	35	38	information	information	NN
3262	35	39	.	.	.
3262	36	1	This	this	DT
3262	36	2	process	process	NN
3262	36	3	is	be	VBZ
3262	36	4	illustrated	illustrate	VBN
3262	36	5	in	in	IN
3262	36	6	figure	figure	NN
3262	36	7	1	1	CD
3262	36	8	.	.	.
3262	37	1	The	the	DT
3262	37	2	following	follow	VBG
3262	37	3	sections	section	NNS
3262	37	4	will	will	MD
3262	37	5	examine	examine	VB
3262	37	6	some	some	DT
3262	37	7	exist-	exist-	NN
3262	37	8	ing	ing	NN
3262	37	9	content	content	NN
3262	37	10	-	-	HYPH
3262	37	11	based	base	VBN
3262	37	12	information	information	NN
3262	37	13	retrieval	retrieval	NN
3262	37	14	techniques	technique	NNS
3262	37	15	for	for	IN
3262	37	16	most	most	JJS
3262	37	17	common	common	JJ
3262	37	18	information	information	NN
3262	37	19	formats	format	NNS
3262	37	20	(	(	-LRB-
3262	37	21	image	image	NN
3262	37	22	,	,	,
3262	37	23	audio	audio	NN
3262	37	24	,	,	,
3262	37	25	and	and	CC
3262	37	26	video	video	NN
3262	37	27	)	)	-RRB-
3262	37	28	in	in	IN
3262	37	29	digital	digital	JJ
3262	37	30	libraries	library	NNS
3262	37	31	,	,	,
3262	37	32	as	as	RB
3262	37	33	well	well	RB
3262	37	34	as	as	IN
3262	37	35	their	-PRON-	PRP$
3262	37	36	limitations	limitation	NNS
3262	37	37	and	and	CC
3262	37	38	trends	trend	NNS
3262	37	39	.	.	.
3262	38	1	Gary	Gary	NNP
3262	38	2	(	(	-LRB-
3262	38	3	Gang	Gang	NNP
3262	38	4	)	)	-RRB-
3262	38	5	Wan	Wan	NNP
3262	38	6	(	(	-LRB-
3262	38	7	gwan@tamu.edu	gwan@tamu.edu	-RRB-
3262	38	8	)	)	-RRB-
3262	38	9	is	be	VBZ
3262	38	10	a	a	DT
3262	38	11	Science	Science	NNP
3262	38	12	Librarian	Librarian	NNP
3262	38	13	and	and	CC
3262	38	14	Assistant	Assistant	NNP
3262	38	15	Professor	Professor	NNP
3262	38	16	,	,	,
3262	38	17	and	and	CC
3262	38	18	zao	zao	NNP
3262	38	19	Liu	Liu	NNP
3262	38	20	(	(	-LRB-
3262	38	21	zliu@tamu.edu	zliu@tamu.edu	NNP
3262	38	22	)	)	-RRB-
3262	38	23	is	be	VBZ
3262	38	24	a	a	DT
3262	38	25	Distance	Distance	NNP
3262	38	26	Learning	Learning	NNP
3262	38	27	Librarian	Librarian	NNP
3262	38	28	and	and	CC
3262	38	29	Assistant	Assistant	NNP
3262	38	30	Professor	Professor	NNP
3262	38	31	at	at	IN
3262	38	32	Sterling	Sterling	NNP
3262	38	33	C.	C.	NNP
3262	38	34	Evans	Evans	NNP
3262	38	35	Library	Library	NNP
3262	38	36	,	,	,
3262	38	37	Texas	Texas	NNP
3262	38	38	A&M	A&M	NNP
3262	38	39	University	University	NNP
3262	38	40	,	,	,
3262	38	41	College	College	NNP
3262	38	42	Station	Station	NNP
3262	38	43	,	,	,
3262	38	44	Texas	Texas	NNP
3262	38	45	.	.	.
3262	39	1	Gary	Gary	NNP
3262	39	2	(	(	-LRB-
3262	39	3	Gang	Gang	NNP
3262	39	4	)	)	-RRB-
3262	39	5	Wan	Wan	NNP
3262	39	6	and	and	CC
3262	39	7	Zao	Zao	NNP
3262	39	8	Liu	Liu	NNP
3262	39	9	42	42	CD
3262	39	10	inFoRmation	information	NN
3262	39	11	tECHnoLoGY	tECHnoLoGY	:
3262	39	12	anD	and	CC
3262	39	13	LiBRaRiEs	LiBRaRiEs	NNP
3262	39	14	|	|	NNP
3262	39	15	maRCH	maRCH	VBZ
3262	39	16	200842	200842	CD
3262	39	17	inFoRmation	information	NN
3262	39	18	tECHnoLoGY	tECHnoLoGY	:
3262	39	19	anD	and	CC
3262	39	20	LiBRaRiEs	LiBRaRiEs	NNP
3262	39	21	|	|	NNP
3262	39	22	maRCH	march	CD
3262	39	23	2008	2008	CD
3262	39	24	■	■	NFP
3262	39	25	Content	content	NN
3262	39	26	-	-	HYPH
3262	39	27	based	base	VBN
3262	39	28	image	image	NN
3262	39	29	retrieval	retrieval	NN
3262	39	30	There	there	EX
3262	39	31	have	have	VBP
3262	39	32	been	be	VBN
3262	39	33	a	a	DT
3262	39	34	large	large	JJ
3262	39	35	number	number	NN
3262	39	36	of	of	IN
3262	39	37	different	different	JJ
3262	39	38	content-	content-	NNP
3262	39	39	based	base	VBN
3262	39	40	image	image	NN
3262	39	41	retrieval	retrieval	NN
3262	39	42	(	(	-LRB-
3262	39	43	CBIR	CBIR	NNP
3262	39	44	)	)	-RRB-
3262	39	45	systems	system	NNS
3262	39	46	proposed	propose	VBN
3262	39	47	in	in	IN
3262	39	48	the	the	DT
3262	39	49	last	last	JJ
3262	39	50	few	few	JJ
3262	39	51	years	year	NNS
3262	39	52	,	,	,
3262	39	53	either	either	CC
3262	39	54	building	build	VBG
3262	39	55	on	on	IN
3262	39	56	prior	prior	JJ
3262	39	57	work	work	NN
3262	39	58	or	or	CC
3262	39	59	exploring	explore	VBG
3262	39	60	novel	novel	JJ
3262	39	61	directions	direction	NNS
3262	39	62	.	.	.
3262	40	1	One	one	CD
3262	40	2	similarity	similarity	NN
3262	40	3	among	among	IN
3262	40	4	these	these	DT
3262	40	5	systems	system	NNS
3262	40	6	is	be	VBZ
3262	40	7	that	that	IN
3262	40	8	most	most	JJS
3262	40	9	perform	perform	VBP
3262	40	10	feature	feature	NN
3262	40	11	extraction	extraction	NN
3262	40	12	as	as	IN
3262	40	13	the	the	DT
3262	40	14	first	first	JJ
3262	40	15	step	step	NN
3262	40	16	in	in	IN
3262	40	17	the	the	DT
3262	40	18	process	process	NN
3262	40	19	,	,	,
3262	40	20	obtaining	obtain	VBG
3262	40	21	global	global	JJ
3262	40	22	image	image	NN
3262	40	23	features	feature	NNS
3262	40	24	such	such	JJ
3262	40	25	as	as	IN
3262	40	26	color	color	NN
3262	40	27	,	,	,
3262	40	28	shape	shape	NN
3262	40	29	,	,	,
3262	40	30	and	and	CC
3262	40	31	texture	texture	NN
3262	40	32	(	(	-LRB-
3262	40	33	Datta	Datta	NNP
3262	40	34	et	et	FW
3262	40	35	al	al	NNP
3262	40	36	.	.	NNP
3262	40	37	,	,	,
3262	40	38	2005	2005	CD
3262	40	39	)	)	-RRB-
3262	40	40	.	.	.
3262	41	1	One	one	CD
3262	41	2	of	of	IN
3262	41	3	the	the	DT
3262	41	4	most	most	RBS
3262	41	5	well	well	RB
3262	41	6	-	-	HYPH
3262	41	7	known	know	VBN
3262	41	8	CBIR	CBIR	NNP
3262	41	9	systems	system	NNS
3262	41	10	is	be	VBZ
3262	41	11	query	query	NN
3262	41	12	by	by	IN
3262	41	13	image	image	NN
3262	41	14	content	content	NN
3262	41	15	(	(	-LRB-
3262	41	16	QBIC	QBIC	NNP
3262	41	17	)	)	-RRB-
3262	41	18	,	,	,
3262	41	19	which	which	WDT
3262	41	20	was	be	VBD
3262	41	21	developed	develop	VBN
3262	41	22	by	by	IN
3262	41	23	IBM	IBM	NNP
3262	41	24	.	.	.
3262	42	1	It	-PRON-	PRP
3262	42	2	uses	use	VBZ
3262	42	3	several	several	JJ
3262	42	4	different	different	JJ
3262	42	5	features	feature	NNS
3262	42	6	,	,	,
3262	42	7	including	include	VBG
3262	42	8	color	color	NN
3262	42	9	,	,	,
3262	42	10	sketches	sketch	NNS
3262	42	11	,	,	,
3262	42	12	texture	texture	NN
3262	42	13	,	,	,
3262	42	14	shape	shape	NN
3262	42	15	,	,	,
3262	42	16	and	and	CC
3262	42	17	example	example	NN
3262	42	18	images	image	NNS
3262	42	19	to	to	TO
3262	42	20	retrieve	retrieve	VB
3262	42	21	images	image	NNS
3262	42	22	from	from	IN
3262	42	23	image	image	NN
3262	42	24	and	and	CC
3262	42	25	video	video	NN
3262	42	26	databases	database	NNS
3262	42	27	.	.	.
3262	43	1	Since	since	IN
3262	43	2	its	-PRON-	PRP$
3262	43	3	launch	launch	NN
3262	43	4	in	in	IN
3262	43	5	1995	1995	CD
3262	43	6	,	,	,
3262	43	7	the	the	DT
3262	43	8	QBIC	QBIC	NNP
3262	43	9	model	model	NN
3262	43	10	has	have	VBZ
3262	43	11	been	be	VBN
3262	43	12	employed	employ	VBN
3262	43	13	for	for	IN
3262	43	14	quite	quite	PDT
3262	43	15	a	a	DT
3262	43	16	few	few	JJ
3262	43	17	digital	digital	JJ
3262	43	18	libraries	library	NNS
3262	43	19	or	or	CC
3262	43	20	collections	collection	NNS
3262	43	21	.	.	.
3262	44	1	One	one	CD
3262	44	2	recent	recent	JJ
3262	44	3	adopter	adopter	NN
3262	44	4	is	be	VBZ
3262	44	5	the	the	DT
3262	44	6	State	State	NNP
3262	44	7	Hermitage	Hermitage	NNP
3262	44	8	Museum	Museum	NNP
3262	44	9	in	in	IN
3262	44	10	Russia	Russia	NNP
3262	44	11	(	(	-LRB-
3262	44	12	www.hermitage	www.hermitage	NNP
3262	44	13	.	.	.
3262	45	1	ru	ru	NNP
3262	45	2	)	)	-RRB-
3262	45	3	,	,	,
3262	45	4	which	which	WDT
3262	45	5	uses	use	VBZ
3262	45	6	QBIC	QBIC	NNP
3262	45	7	for	for	IN
3262	45	8	its	-PRON-	PRP$
3262	45	9	Web	web	NN
3262	45	10	-	-	HYPH
3262	45	11	based	base	VBN
3262	45	12	digital	digital	JJ
3262	45	13	image	image	NN
3262	45	14	col-	col-	NN
3262	45	15	lection	lection	NN
3262	45	16	.	.	.
3262	46	1	Users	user	NNS
3262	46	2	can	can	MD
3262	46	3	find	find	VB
3262	46	4	artwork	artwork	NN
3262	46	5	images	image	NNS
3262	46	6	by	by	IN
3262	46	7	selecting	select	VBG
3262	46	8	colors	color	NNS
3262	46	9	from	from	IN
3262	46	10	a	a	DT
3262	46	11	palette	palette	NN
3262	46	12	or	or	CC
3262	46	13	by	by	IN
3262	46	14	sketching	sketch	VBG
3262	46	15	shapes	shape	NNS
3262	46	16	on	on	IN
3262	46	17	a	a	DT
3262	46	18	canvas	canvas	NN
3262	46	19	.	.	.
3262	47	1	The	the	DT
3262	47	2	user	user	NN
3262	47	3	can	can	MD
3262	47	4	also	also	RB
3262	47	5	refine	refine	VB
3262	47	6	existing	exist	VBG
3262	47	7	search	search	NN
3262	47	8	results	result	NNS
3262	47	9	by	by	IN
3262	47	10	requesting	request	VBG
3262	47	11	all	all	DT
3262	47	12	artwork	artwork	NN
3262	47	13	images	image	NNS
3262	47	14	with	with	IN
3262	47	15	similar	similar	JJ
3262	47	16	visual	visual	JJ
3262	47	17	attributes	attribute	NNS
3262	47	18	.	.	.
3262	48	1	The	the	DT
3262	48	2	following	follow	VBG
3262	48	3	screenshots	screenshot	NNS
3262	48	4	demonstrate	demonstrate	VBP
3262	48	5	how	how	WRB
3262	48	6	a	a	DT
3262	48	7	user	user	NN
3262	48	8	can	can	MD
3262	48	9	do	do	VB
3262	48	10	a	a	DT
3262	48	11	content	content	NN
3262	48	12	-	-	HYPH
3262	48	13	based	base	VBN
3262	48	14	image	image	NN
3262	48	15	search	search	NN
3262	48	16	with	with	IN
3262	48	17	QBIC	QBIC	NNP
3262	48	18	technology	technology	NN
3262	48	19	.	.	.
3262	49	1	In	in	IN
3262	49	2	figure	figure	NN
3262	49	3	2.1	2.1	CD
3262	49	4	,	,	,
3262	49	5	the	the	DT
3262	49	6	user	user	NN
3262	49	7	chooses	choose	VBZ
3262	49	8	a	a	DT
3262	49	9	color	color	NN
3262	49	10	from	from	IN
3262	49	11	the	the	DT
3262	49	12	palette	palette	NN
3262	49	13	and	and	CC
3262	49	14	composes	compose	VBZ
3262	49	15	the	the	DT
3262	49	16	color	color	NN
3262	49	17	schema	schema	NN
3262	49	18	of	of	IN
3262	49	19	artwork	artwork	NN
3262	49	20	he	-PRON-	PRP
3262	49	21	or	or	CC
3262	49	22	she	-PRON-	PRP
3262	49	23	is	be	VBZ
3262	49	24	looking	look	VBG
3262	49	25	for	for	IN
3262	49	26	.	.	.
3262	50	1	Figure	figure	NN
3262	50	2	2.2	2.2	CD
3262	50	3	shows	show	VBZ
3262	50	4	the	the	DT
3262	50	5	artwork	artwork	NN
3262	50	6	images	image	NNS
3262	50	7	that	that	WDT
3262	50	8	match	match	VBP
3262	50	9	the	the	DT
3262	50	10	query	query	NN
3262	50	11	schema	schema	NNP
3262	50	12	.	.	.
3262	51	1	Another	another	DT
3262	51	2	example	example	NN
3262	51	3	of	of	IN
3262	51	4	digital	digital	JJ
3262	51	5	libraries	library	NNS
3262	51	6	or	or	CC
3262	51	7	collections	collection	NNS
3262	51	8	that	that	WDT
3262	51	9	have	have	VBP
3262	51	10	incorporated	incorporate	VBN
3262	51	11	CBIR	CBIR	NNP
3262	51	12	technology	technology	NN
3262	51	13	is	be	VBZ
3262	51	14	the	the	DT
3262	51	15	National	National	NNP
3262	51	16	Science	Science	NNP
3262	51	17	Foundation	Foundation	NNP
3262	51	18	’s	’s	POS
3262	51	19	International	International	NNP
3262	51	20	Digital	Digital	NNP
3262	51	21	Library	Library	NNP
3262	51	22	Project	Project	NNP
3262	51	23	(	(	-LRB-
3262	51	24	www.memorynet.org	www.memorynet.org	NNP
3262	51	25	)	)	-RRB-
3262	51	26	,	,	,
3262	51	27	a	a	DT
3262	51	28	project	project	NN
3262	51	29	that	that	WDT
3262	51	30	is	be	VBZ
3262	51	31	composed	compose	VBN
3262	51	32	of	of	IN
3262	51	33	several	several	JJ
3262	51	34	image	image	NN
3262	51	35	collections	collection	NNS
3262	51	36	.	.	.
3262	52	1	The	the	DT
3262	52	2	information	information	NN
3262	52	3	retrieval	retrieval	NN
3262	52	4	sys-	sys-	NNP
3262	52	5	tem	tem	NN
3262	52	6	for	for	IN
3262	52	7	these	these	DT
3262	52	8	collections	collection	NNS
3262	52	9	includes	include	VBZ
3262	52	10	both	both	CC
3262	52	11	a	a	DT
3262	52	12	traditional	traditional	JJ
3262	52	13	text	text	NN
3262	52	14	-	-	HYPH
3262	52	15	based	base	VBN
3262	52	16	search	search	NN
3262	52	17	engine	engine	NN
3262	52	18	and	and	CC
3262	52	19	a	a	DT
3262	52	20	CBIR	CBIR	NNP
3262	52	21	system	system	NN
3262	52	22	called	call	VBN
3262	52	23	SIMPLIcity	simplicity	NN
3262	52	24	(	(	-LRB-
3262	52	25	Semantics	Semantics	NNP
3262	52	26	-	-	HYPH
3262	52	27	sensitive	sensitive	JJ
3262	52	28	Integrated	Integrated	NNP
3262	52	29	Matching	Matching	NNP
3262	52	30	for	for	IN
3262	52	31	Picture	Picture	NNP
3262	52	32	Libraries	Libraries	NNPS
3262	52	33	)	)	-RRB-
3262	52	34	developed	develop	VBN
3262	52	35	by	by	IN
3262	52	36	Wang	Wang	NNP
3262	52	37	et	et	NNP
3262	52	38	al	al	NNP
3262	52	39	.	.	.
3262	53	1	(	(	-LRB-
3262	53	2	2001	2001	CD
3262	53	3	)	)	-RRB-
3262	53	4	of	of	IN
3262	53	5	Pennsylvania	Pennsylvania	NNP
3262	53	6	State	State	NNP
3262	53	7	University	University	NNP
3262	53	8	.	.	.
3262	54	1	From	from	IN
3262	54	2	the	the	DT
3262	54	3	front	front	JJ
3262	54	4	page	page	NN
3262	54	5	of	of	IN
3262	54	6	these	these	DT
3262	54	7	image	image	NN
3262	54	8	collections	collection	NNS
3262	54	9	,	,	,
3262	54	10	a	a	DT
3262	54	11	user	user	NN
3262	54	12	can	can	MD
3262	54	13	choose	choose	VB
3262	54	14	to	to	TO
3262	54	15	display	display	VB
3262	54	16	a	a	DT
3262	54	17	random	random	JJ
3262	54	18	group	group	NN
3262	54	19	of	of	IN
3262	54	20	images	image	NNS
3262	54	21	(	(	-LRB-
3262	54	22	figure	figure	VB
3262	54	23	3.1	3.1	CD
3262	54	24	)	)	-RRB-
3262	54	25	.	.	.
3262	55	1	Below	below	IN
3262	55	2	each	each	DT
3262	55	3	image	image	NN
3262	55	4	is	be	VBZ
3262	55	5	a	a	DT
3262	55	6	“	"	``
3262	55	7	similar	similar	JJ
3262	55	8	”	"	''
3262	55	9	button	button	NN
3262	55	10	;	;	,
3262	55	11	clicking	click	VBG
3262	55	12	this	this	DT
3262	55	13	allows	allow	VBZ
3262	55	14	the	the	DT
3262	55	15	user	user	NN
3262	55	16	to	to	TO
3262	55	17	view	view	VB
3262	55	18	a	a	DT
3262	55	19	group	group	NN
3262	55	20	of	of	IN
3262	55	21	images	image	NNS
3262	55	22	that	that	WDT
3262	55	23	contain	contain	VBP
3262	55	24	similar	similar	JJ
3262	55	25	objects	object	NNS
3262	55	26	to	to	IN
3262	55	27	the	the	DT
3262	55	28	previously	previously	RB
3262	55	29	selected	select	VBN
3262	55	30	one	one	CD
3262	55	31	(	(	-LRB-
3262	55	32	figure	figure	NN
3262	55	33	3.2	3.2	CD
3262	55	34	)	)	-RRB-
3262	55	35	.	.	.
3262	56	1	By	by	IN
3262	56	2	providing	provide	VBG
3262	56	3	feedback	feedback	NN
3262	56	4	to	to	IN
3262	56	5	the	the	DT
3262	56	6	search	search	NN
3262	56	7	engine	engine	NN
3262	56	8	this	this	DT
3262	56	9	way	way	NN
3262	56	10	,	,	,
3262	56	11	the	the	DT
3262	56	12	user	user	NN
3262	56	13	can	can	MD
3262	56	14	find	find	VB
3262	56	15	images	image	NNS
3262	56	16	of	of	IN
3262	56	17	desired	desire	VBN
3262	56	18	objects	object	NNS
3262	56	19	without	without	IN
3262	56	20	knowing	know	VBG
3262	56	21	their	-PRON-	PRP$
3262	56	22	names	name	NNS
3262	56	23	or	or	CC
3262	56	24	descriptions	description	NNS
3262	56	25	.	.	.
3262	57	1	Simply	simply	RB
3262	57	2	put	put	VBN
3262	57	3	,	,	,
3262	57	4	SIMPLIcity	simplicity	NN
3262	57	5	segments	segment	NNS
3262	57	6	each	each	DT
3262	57	7	image	image	NN
3262	57	8	into	into	IN
3262	57	9	small	small	JJ
3262	57	10	regions	region	NNS
3262	57	11	,	,	,
3262	57	12	extracts	extract	VBZ
3262	57	13	several	several	JJ
3262	57	14	features	feature	NNS
3262	57	15	(	(	-LRB-
3262	57	16	such	such	JJ
3262	57	17	as	as	IN
3262	57	18	color	color	NN
3262	57	19	,	,	,
3262	57	20	Figure	figure	NN
3262	57	21	1	1	CD
3262	57	22	.	.	.
3262	58	1	The	the	DT
3262	58	2	general	general	JJ
3262	58	3	process	process	NN
3262	58	4	of	of	IN
3262	58	5	content	content	NN
3262	58	6	-	-	HYPH
3262	58	7	based	base	VBN
3262	58	8	information	information	NN
3262	58	9	retrieval	retrieval	NN
3262	58	10	Figure	Figure	NNP
3262	58	11	2.1	2.1	CD
3262	58	12	.	.	.
3262	59	1	A	a	DT
3262	59	2	user	user	NN
3262	59	3	query	query	NN
3262	59	4	Figure	figure	NN
3262	59	5	2.2	2.2	CD
3262	59	6	.	.	.
3262	60	1	The	the	DT
3262	60	2	search	search	NN
3262	60	3	results	result	VBZ
3262	60	4	for	for	IN
3262	60	5	this	this	DT
3262	60	6	query	query	NN
3262	60	7	aRtiCLE	article	CD
3262	60	8	titLE	title	CD
3262	60	9	|	|	NNP
3262	60	10	autHoR	autHoR	NNP
3262	60	11	43ContEnt	43ContEnt	NNP
3262	60	12	-	-	HYPH
3262	60	13	BasED	BasED	NNP
3262	60	14	inFoRmation	information	NN
3262	60	15	REtRiEvaL	REtRiEvaL	NNP
3262	60	16	anD	and	CC
3262	60	17	DiGitaL	DiGitaL	NNP
3262	60	18	LiBRaRiEs	LiBRaRiEs	NNP
3262	60	19	|	|	NNP
3262	60	20	Wan	Wan	NNP
3262	60	21	anD	and	CC
3262	60	22	Liu	Liu	NNP
3262	60	23	43	43	CD
3262	60	24	location	location	NN
3262	60	25	,	,	,
3262	60	26	and	and	CC
3262	60	27	shape	shape	NN
3262	60	28	)	)	-RRB-
3262	60	29	from	from	IN
3262	60	30	these	these	DT
3262	60	31	small	small	JJ
3262	60	32	regions	region	NNS
3262	60	33	,	,	,
3262	60	34	and	and	CC
3262	60	35	clas-	clas-	NNP
3262	60	36	sifies	sifie	VBZ
3262	60	37	these	these	DT
3262	60	38	regions	region	NNS
3262	60	39	into	into	IN
3262	60	40	some	some	DT
3262	60	41	semantic	semantic	JJ
3262	60	42	categories	category	NNS
3262	60	43	(	(	-LRB-
3262	60	44	such	such	JJ
3262	60	45	as	as	IN
3262	60	46	textured	texture	VBN
3262	60	47	/	/	SYM
3262	60	48	nontextured	nontextured	JJ
3262	60	49	and	and	CC
3262	60	50	graph	graph	NNP
3262	60	51	/	/	SYM
3262	60	52	photograph	photograph	NNP
3262	60	53	)	)	-RRB-
3262	60	54	.	.	.
3262	61	1	When	when	WRB
3262	61	2	computing	compute	VBG
3262	61	3	the	the	DT
3262	61	4	similarity	similarity	NN
3262	61	5	between	between	IN
3262	61	6	the	the	DT
3262	61	7	query	query	NN
3262	61	8	image	image	NN
3262	61	9	and	and	CC
3262	61	10	images	image	NNS
3262	61	11	in	in	IN
3262	61	12	the	the	DT
3262	61	13	database	database	NN
3262	61	14	,	,	,
3262	61	15	all	all	PDT
3262	61	16	these	these	DT
3262	61	17	features	feature	NNS
3262	61	18	will	will	MD
3262	61	19	be	be	VB
3262	61	20	consid-	consid-	NN
3262	61	21	ered	ered	JJ
3262	61	22	and	and	CC
3262	61	23	integrated	integrate	VBN
3262	61	24	,	,	,
3262	61	25	and	and	CC
3262	61	26	best	good	JJS
3262	61	27	matching	matching	NN
3262	61	28	results	result	NNS
3262	61	29	will	will	MD
3262	61	30	be	be	VB
3262	61	31	retrieved	retrieve	VBN
3262	61	32	(	(	-LRB-
3262	61	33	Wang	Wang	NNP
3262	61	34	et	et	NNP
3262	61	35	al	al	NNP
3262	61	36	.	.	NNP
3262	61	37	,	,	,
3262	61	38	2001	2001	CD
3262	61	39	)	)	-RRB-
3262	61	40	.	.	.
3262	62	1	Similar	similar	JJ
3262	62	2	applications	application	NNS
3262	62	3	of	of	IN
3262	62	4	CBIR	CBIR	NNP
3262	62	5	technology	technology	NN
3262	62	6	in	in	IN
3262	62	7	digital	digital	JJ
3262	62	8	libraries	library	NNS
3262	62	9	include	include	VBP
3262	62	10	the	the	DT
3262	62	11	University	University	NNP
3262	62	12	of	of	IN
3262	62	13	California	California	NNP
3262	62	14	–	–	:
3262	62	15	Berkeley	Berkeley	NNP
3262	62	16	’s	’s	POS
3262	62	17	Digital	Digital	NNP
3262	62	18	Library	Library	NNP
3262	62	19	Project	Project	NNP
3262	62	20	(	(	-LRB-
3262	62	21	http://bnhm.berkeley.edu	http://bnhm.berkeley.edu	NNP
3262	62	22	)	)	-RRB-
3262	62	23	,	,	,
3262	62	24	the	the	DT
3262	62	25	National	National	NNP
3262	62	26	STEM	STEM	NNP
3262	62	27	Digital	Digital	NNP
3262	62	28	Library	Library	NNP
3262	62	29	(	(	-LRB-
3262	62	30	ongoing	ongoing	JJ
3262	62	31	)	)	-RRB-
3262	62	32	,	,	,
3262	62	33	and	and	CC
3262	62	34	Virginia	Virginia	NNP
3262	62	35	Tech	Tech	NNP
3262	62	36	’s	’s	POS
3262	62	37	anthropology	anthropology	NN
3262	62	38	digital	digital	JJ
3262	62	39	library	library	NN
3262	62	40	,	,	,
3262	62	41	ETANA	ETANA	NNP
3262	62	42	(	(	-LRB-
3262	62	43	ongoing	ongoing	JJ
3262	62	44	)	)	-RRB-
3262	62	45	.	.	.
3262	63	1	While	while	IN
3262	63	2	these	these	DT
3262	63	3	feature	feature	NN
3262	63	4	-	-	HYPH
3262	63	5	based	base	VBN
3262	63	6	approaches	approach	NNS
3262	63	7	have	have	VBP
3262	63	8	been	be	VBN
3262	63	9	explored	explore	VBN
3262	63	10	over	over	IN
3262	63	11	the	the	DT
3262	63	12	years	year	NNS
3262	63	13	,	,	,
3262	63	14	an	an	DT
3262	63	15	emerging	emerge	VBG
3262	63	16	new	new	JJ
3262	63	17	research	research	NN
3262	63	18	direction	direction	NN
3262	63	19	in	in	IN
3262	63	20	CBIR	CBIR	NNP
3262	63	21	is	be	VBZ
3262	63	22	automatic	automatic	JJ
3262	63	23	concept	concept	NN
3262	63	24	recognition	recognition	NN
3262	63	25	and	and	CC
3262	63	26	annotation	annotation	NN
3262	63	27	.	.	.
3262	64	1	Ideally	ideally	RB
3262	64	2	,	,	,
3262	64	3	automatic	automatic	JJ
3262	64	4	concept	concept	NN
3262	64	5	recognition	recognition	NN
3262	64	6	and	and	CC
3262	64	7	annotation	annotation	NN
3262	64	8	can	can	MD
3262	64	9	discover	discover	VB
3262	64	10	the	the	DT
3262	64	11	concepts	concept	NNS
3262	64	12	that	that	WDT
3262	64	13	an	an	DT
3262	64	14	image	image	NN
3262	64	15	con-	con-	NN
3262	64	16	veys	vey	NNS
3262	64	17	and	and	CC
3262	64	18	assign	assign	VB
3262	64	19	a	a	DT
3262	64	20	set	set	NN
3262	64	21	of	of	IN
3262	64	22	metadata	metadata	NN
3262	64	23	to	to	IN
3262	64	24	it	-PRON-	PRP
3262	64	25	,	,	,
3262	64	26	thus	thus	RB
3262	64	27	allowing	allow	VBG
3262	64	28	image	image	NN
3262	64	29	search	search	NN
3262	64	30	through	through	IN
3262	64	31	the	the	DT
3262	64	32	use	use	NN
3262	64	33	of	of	IN
3262	64	34	text	text	NN
3262	64	35	.	.	.
3262	65	1	A	a	DT
3262	65	2	trusted	trust	VBN
3262	65	3	automatic	automatic	JJ
3262	65	4	concept	concept	NN
3262	65	5	recognition	recognition	NN
3262	65	6	and	and	CC
3262	65	7	annotation	annotation	NN
3262	65	8	system	system	NN
3262	65	9	can	can	MD
3262	65	10	be	be	VB
3262	65	11	a	a	DT
3262	65	12	good	good	JJ
3262	65	13	solution	solution	NN
3262	65	14	for	for	IN
3262	65	15	large	large	JJ
3262	65	16	data	data	NN
3262	65	17	sets	set	NNS
3262	65	18	.	.	.
3262	66	1	However	however	RB
3262	66	2	,	,	,
3262	66	3	the	the	DT
3262	66	4	semantic	semantic	JJ
3262	66	5	gap	gap	NN
3262	66	6	between	between	IN
3262	66	7	computer	computer	NN
3262	66	8	processors	processor	NNS
3262	66	9	and	and	CC
3262	66	10	human	human	JJ
3262	66	11	brains	brain	NNS
3262	66	12	remains	remain	VBZ
3262	66	13	the	the	DT
3262	66	14	major	major	JJ
3262	66	15	challenge	challenge	NN
3262	66	16	in	in	IN
3262	66	17	the	the	DT
3262	66	18	development	development	NN
3262	66	19	of	of	IN
3262	66	20	a	a	DT
3262	66	21	robust	robust	JJ
3262	66	22	auto-	auto-	NNP
3262	66	23	matic	matic	JJ
3262	66	24	concept	concept	NN
3262	66	25	recognition	recognition	NN
3262	66	26	and	and	CC
3262	66	27	annotation	annotation	NN
3262	66	28	system	system	NN
3262	66	29	(	(	-LRB-
3262	66	30	Datta	Datta	NNP
3262	66	31	et	et	FW
3262	66	32	al	al	NNP
3262	66	33	.	.	NNP
3262	66	34	,	,	,
3262	66	35	2005	2005	CD
3262	66	36	)	)	-RRB-
3262	66	37	.	.	.
3262	67	1	A	a	DT
3262	67	2	recent	recent	JJ
3262	67	3	example	example	NN
3262	67	4	of	of	IN
3262	67	5	efforts	effort	NNS
3262	67	6	in	in	IN
3262	67	7	this	this	DT
3262	67	8	field	field	NN
3262	67	9	is	be	VBZ
3262	67	10	Li	Li	NNP
3262	67	11	and	and	CC
3262	67	12	Wang	Wang	NNP
3262	67	13	’s	’s	NNP
3262	67	14	ALIPR	ALIPR	NNP
3262	67	15	(	(	-LRB-
3262	67	16	Automatic	Automatic	NNP
3262	67	17	Linguistic	Linguistic	NNP
3262	67	18	Indexing	Indexing	NNP
3262	67	19	of	of	IN
3262	67	20	Pictures	Pictures	NNPS
3262	67	21	—	—	:
3262	67	22	Real	real	JJ
3262	67	23	Time	time	NN
3262	67	24	,	,	,
3262	67	25	http://alipr.com	http://alipr.com	ADD
3262	67	26	)	)	-RRB-
3262	67	27	project	project	NN
3262	67	28	(	(	-LRB-
3262	67	29	2006	2006	CD
3262	67	30	)	)	-RRB-
3262	67	31	.	.	.
3262	68	1	Through	through	IN
3262	68	2	a	a	DT
3262	68	3	Web	web	NN
3262	68	4	interface	interface	NN
3262	68	5	,	,	,
3262	68	6	users	user	NNS
3262	68	7	are	be	VBP
3262	68	8	able	able	JJ
3262	68	9	to	to	TO
3262	68	10	search	search	VB
3262	68	11	images	image	NNS
3262	68	12	in	in	IN
3262	68	13	several	several	JJ
3262	68	14	different	different	JJ
3262	68	15	ways	way	NNS
3262	68	16	:	:	:
3262	68	17	They	-PRON-	PRP
3262	68	18	may	may	MD
3262	68	19	do	do	VB
3262	68	20	text	text	NN
3262	68	21	searches	search	NNS
3262	68	22	and	and	CC
3262	68	23	provide	provide	VB
3262	68	24	feedback	feedback	NN
3262	68	25	to	to	IN
3262	68	26	the	the	DT
3262	68	27	system	system	NN
3262	68	28	to	to	TO
3262	68	29	find	find	VB
3262	68	30	similar	similar	JJ
3262	68	31	images	image	NNS
3262	68	32	.	.	.
3262	69	1	Users	user	NNS
3262	69	2	may	may	MD
3262	69	3	also	also	RB
3262	69	4	upload	upload	VB
3262	69	5	an	an	DT
3262	69	6	image	image	NN
3262	69	7	,	,	,
3262	69	8	and	and	CC
3262	69	9	the	the	DT
3262	69	10	system	system	NN
3262	69	11	will	will	MD
3262	69	12	per-	per-	RB
3262	69	13	form	form	VB
3262	69	14	concept	concept	NN
3262	69	15	analysis	analysis	NN
3262	69	16	and	and	CC
3262	69	17	generate	generate	VB
3262	69	18	a	a	DT
3262	69	19	set	set	NN
3262	69	20	of	of	IN
3262	69	21	annotations	annotation	NNS
3262	69	22	or	or	CC
3262	69	23	tags	tag	NNS
3262	69	24	automatically	automatically	RB
3262	69	25	,	,	,
3262	69	26	as	as	IN
3262	69	27	shown	show	VBN
3262	69	28	in	in	IN
3262	69	29	figure	figure	NN
3262	69	30	4	4	CD
3262	69	31	.	.	.
3262	70	1	The	the	DT
3262	70	2	system	system	NN
3262	70	3	then	then	RB
3262	70	4	retrieves	retrieve	VBZ
3262	70	5	images	image	NNS
3262	70	6	from	from	IN
3262	70	7	the	the	DT
3262	70	8	database	database	NN
3262	70	9	that	that	WDT
3262	70	10	are	be	VBP
3262	70	11	visually	visually	RB
3262	70	12	similar	similar	JJ
3262	70	13	to	to	IN
3262	70	14	the	the	DT
3262	70	15	uploaded	uploaded	JJ
3262	70	16	image	image	NN
3262	70	17	.	.	.
3262	71	1	In	in	IN
3262	71	2	the	the	DT
3262	71	3	process	process	NN
3262	71	4	of	of	IN
3262	71	5	auto-	auto-	FW
3262	71	6	matic	matic	JJ
3262	71	7	annotation	annotation	NN
3262	71	8	,	,	,
3262	71	9	if	if	IN
3262	71	10	the	the	DT
3262	71	11	user	user	NN
3262	71	12	does	do	VBZ
3262	71	13	n’t	not	RB
3262	71	14	think	think	VB
3262	71	15	the	the	DT
3262	71	16	tags	tag	NNS
3262	71	17	given	give	VBN
3262	71	18	by	by	IN
3262	71	19	the	the	DT
3262	71	20	system	system	NN
3262	71	21	are	be	VBP
3262	71	22	suitable	suitable	JJ
3262	71	23	,	,	,
3262	71	24	he	-PRON-	PRP
3262	71	25	or	or	CC
3262	71	26	she	-PRON-	PRP
3262	71	27	can	can	MD
3262	71	28	input	input	VB
3262	71	29	other	other	JJ
3262	71	30	tags	tag	NNS
3262	71	31	to	to	TO
3262	71	32	describe	describe	VB
3262	71	33	the	the	DT
3262	71	34	image	image	NN
3262	71	35	.	.	.
3262	72	1	This	this	DT
3262	72	2	is	be	VBZ
3262	72	3	also	also	RB
3262	72	4	the	the	DT
3262	72	5	“	"	``
3262	72	6	training	training	NN
3262	72	7	”	"	''
3262	72	8	process	process	NN
3262	72	9	for	for	IN
3262	72	10	the	the	DT
3262	72	11	ALIPR	ALIPR	NNP
3262	72	12	system	system	NN
3262	72	13	.	.	.
3262	73	1	Since	since	IN
3262	73	2	CBIR	CBIR	NNP
3262	73	3	is	be	VBZ
3262	73	4	the	the	DT
3262	73	5	major	major	JJ
3262	73	6	research	research	NN
3262	73	7	area	area	NN
3262	73	8	and	and	CC
3262	73	9	has	have	VBZ
3262	73	10	the	the	DT
3262	73	11	lon-	lon-	NN
3262	73	12	gest	g	JJS
3262	73	13	history	history	NN
3262	73	14	in	in	IN
3262	73	15	content	content	NN
3262	73	16	-	-	HYPH
3262	73	17	based	base	VBN
3262	73	18	information	information	NN
3262	73	19	retrieval	retrieval	NN
3262	73	20	,	,	,
3262	73	21	there	there	EX
3262	73	22	are	be	VBP
3262	73	23	many	many	JJ
3262	73	24	models	model	NNS
3262	73	25	,	,	,
3262	73	26	products	product	NNS
3262	73	27	,	,	,
3262	73	28	and	and	CC
3262	73	29	ongoing	ongoing	JJ
3262	73	30	projects	project	NNS
3262	73	31	in	in	IN
3262	73	32	addi-	addi-	JJ
3262	73	33	tion	tion	NN
3262	73	34	to	to	IN
3262	73	35	the	the	DT
3262	73	36	above	above	JJ
3262	73	37	examples	example	NNS
3262	73	38	.	.	.
3262	74	1	As	as	IN
3262	74	2	image	image	NN
3262	74	3	collections	collection	NNS
3262	74	4	become	become	VBP
3262	74	5	a	a	DT
3262	74	6	significant	significant	JJ
3262	74	7	part	part	NN
3262	74	8	of	of	IN
3262	74	9	digital	digital	JJ
3262	74	10	libraries	library	NNS
3262	74	11	,	,	,
3262	74	12	more	more	JJR
3262	74	13	attention	attention	NN
3262	74	14	has	have	VBZ
3262	74	15	been	be	VBN
3262	74	16	paid	pay	VBN
3262	74	17	to	to	IN
3262	74	18	possibilities	possibility	NNS
3262	74	19	of	of	IN
3262	74	20	providing	provide	VBG
3262	74	21	content	content	NN
3262	74	22	-	-	HYPH
3262	74	23	based	base	VBN
3262	74	24	image	image	NN
3262	74	25	search	search	NN
3262	74	26	as	as	IN
3262	74	27	a	a	DT
3262	74	28	complement	complement	NN
3262	74	29	to	to	IN
3262	74	30	existing	exist	VBG
3262	74	31	metadata	metadata	NN
3262	74	32	search	search	NN
3262	74	33	.	.	.
3262	75	1	■	■	NFP
3262	75	2	Content	content	NN
3262	75	3	-	-	HYPH
3262	75	4	based	base	VBN
3262	75	5	audio	audio	JJ
3262	75	6	retrieval	retrieval	NN
3262	75	7	Compared	compare	VBN
3262	75	8	with	with	IN
3262	75	9	CBIR	CBIR	NNP
3262	75	10	,	,	,
3262	75	11	content	content	NN
3262	75	12	-	-	HYPH
3262	75	13	based	base	VBN
3262	75	14	audio	audio	JJ
3262	75	15	retrieval	retrieval	NN
3262	75	16	(	(	-LRB-
3262	75	17	CBAR	CBAR	NNP
3262	75	18	)	)	-RRB-
3262	75	19	is	be	VBZ
3262	75	20	relatively	relatively	RB
3262	75	21	new	new	JJ
3262	75	22	,	,	,
3262	75	23	and	and	CC
3262	75	24	fewer	few	JJR
3262	75	25	research	research	NN
3262	75	26	projects	project	NNS
3262	75	27	on	on	IN
3262	75	28	it	-PRON-	PRP
3262	75	29	can	can	MD
3262	75	30	be	be	VB
3262	75	31	found	find	VBN
3262	75	32	.	.	.
3262	76	1	In	in	IN
3262	76	2	general	general	JJ
3262	76	3	,	,	,
3262	76	4	existing	exist	VBG
3262	76	5	CBAR	CBAR	NNP
3262	76	6	approaches	approach	NNS
3262	76	7	start	start	VBP
3262	76	8	from	from	IN
3262	76	9	the	the	DT
3262	76	10	content	content	NN
3262	76	11	analysis	analysis	NN
3262	76	12	of	of	IN
3262	76	13	audio	audio	JJ
3262	76	14	clips	clip	NNS
3262	76	15	.	.	.
3262	77	1	An	an	DT
3262	77	2	example	example	NN
3262	77	3	of	of	IN
3262	77	4	this	this	DT
3262	77	5	content	content	NN
3262	77	6	analysis	analysis	NN
3262	77	7	is	be	VBZ
3262	77	8	extracting	extract	VBG
3262	77	9	basic	basic	JJ
3262	77	10	audio	audio	JJ
3262	77	11	elements	element	NNS
3262	77	12	,	,	,
3262	77	13	such	such	JJ
3262	77	14	as	as	IN
3262	77	15	duration	duration	NN
3262	77	16	,	,	,
3262	77	17	pitch	pitch	NN
3262	77	18	,	,	,
3262	77	19	amplitude	amplitude	NN
3262	77	20	,	,	,
3262	77	21	brightness	brightness	NN
3262	77	22	,	,	,
3262	77	23	and	and	CC
3262	77	24	band-	band-	VBZ
3262	77	25	Figure	Figure	NNP
3262	77	26	3.1	3.1	CD
3262	77	27	.	.	.
3262	78	1	A	a	DT
3262	78	2	group	group	NN
3262	78	3	of	of	IN
3262	78	4	random	random	JJ
3262	78	5	images	image	NNS
3262	78	6	in	in	IN
3262	78	7	the	the	DT
3262	78	8	collection	collection	NN
3262	78	9	Figure	Figure	NNP
3262	78	10	3.2	3.2	CD
3262	78	11	.	.	.
3262	79	1	CBIR	CBIR	NNP
3262	79	2	results	result	VBZ
3262	79	3	Figure	figure	NN
3262	79	4	4	4	CD
3262	79	5	.	.	.
3262	80	1	ALIPR	ALIPR	NNP
3262	80	2	’s	’s	POS
3262	80	3	automatic	automatic	JJ
3262	80	4	annotation	annotation	NN
3262	80	5	feature	feature	NN
3262	80	6	44	44	CD
3262	80	7	inFoRmation	information	NN
3262	80	8	tECHnoLoGY	tECHnoLoGY	:
3262	80	9	anD	and	CC
3262	80	10	LiBRaRiEs	LiBRaRiEs	NNP
3262	80	11	|	|	NNP
3262	80	12	maRCH	maRCH	VBZ
3262	80	13	200844	200844	CD
3262	80	14	inFoRmation	information	NN
3262	80	15	tECHnoLoGY	tECHnoLoGY	:
3262	80	16	anD	and	CC
3262	80	17	LiBRaRiEs	LiBRaRiEs	NNP
3262	80	18	|	|	NNP
3262	80	19	maRCH	maRCH	NNP
3262	80	20	2008	2008	CD
3262	80	21	width	width	NN
3262	80	22	(	(	-LRB-
3262	80	23	Wold	Wold	NNP
3262	80	24	et	et	NNP
3262	80	25	al	al	NNP
3262	80	26	.	.	NNP
3262	80	27	,	,	,
3262	80	28	1996	1996	CD
3262	80	29	)	)	-RRB-
3262	80	30	.	.	.
3262	81	1	Because	because	IN
3262	81	2	of	of	IN
3262	81	3	the	the	DT
3262	81	4	great	great	JJ
3262	81	5	difficulties	difficulty	NNS
3262	81	6	in	in	IN
3262	81	7	recognizing	recognize	VBG
3262	81	8	audio	audio	JJ
3262	81	9	content	content	NN
3262	81	10	,	,	,
3262	81	11	research	research	NN
3262	81	12	in	in	IN
3262	81	13	this	this	DT
3262	81	14	area	area	NN
3262	81	15	is	be	VBZ
3262	81	16	less	less	RBR
3262	81	17	mature	mature	JJ
3262	81	18	than	than	IN
3262	81	19	that	that	DT
3262	81	20	in	in	IN
3262	81	21	content	content	NN
3262	81	22	-	-	HYPH
3262	81	23	based	base	VBN
3262	81	24	image	image	NN
3262	81	25	and	and	CC
3262	81	26	video	video	NN
3262	81	27	retrieval	retrieval	NN
3262	81	28	.	.	.
3262	82	1	Although	although	IN
3262	82	2	no	no	DT
3262	82	3	CBAR	CBAR	NNP
3262	82	4	system	system	NN
3262	82	5	has	have	VBZ
3262	82	6	been	be	VBN
3262	82	7	found	find	VBN
3262	82	8	to	to	TO
3262	82	9	be	be	VB
3262	82	10	implemented	implement	VBN
3262	82	11	by	by	IN
3262	82	12	any	any	DT
3262	82	13	digital	digital	JJ
3262	82	14	library	library	NN
3262	82	15	so	so	RB
3262	82	16	far	far	RB
3262	82	17	,	,	,
3262	82	18	quite	quite	PDT
3262	82	19	a	a	DT
3262	82	20	few	few	JJ
3262	82	21	projects	project	NNS
3262	82	22	provide	provide	VBP
3262	82	23	good	good	JJ
3262	82	24	prototypes	prototype	NNS
3262	82	25	or	or	CC
3262	82	26	directions	direction	NNS
3262	82	27	.	.	.
3262	83	1	One	one	CD
3262	83	2	good	good	JJ
3262	83	3	example	example	NN
3262	83	4	is	be	VBZ
3262	83	5	Zhang	Zhang	NNP
3262	83	6	and	and	CC
3262	83	7	Kuo	Kuo	NNP
3262	83	8	’s	’s	,
3262	83	9	(	(	-LRB-
3262	83	10	2001	2001	CD
3262	83	11	)	)	-RRB-
3262	83	12	research	research	NN
3262	83	13	project	project	NN
3262	83	14	on	on	IN
3262	83	15	audio	audio	JJ
3262	83	16	classification	classification	NN
3262	83	17	and	and	CC
3262	83	18	retrieval	retrieval	NN
3262	83	19	.	.	.
3262	84	1	The	the	DT
3262	84	2	proto-	proto-	NN
3262	84	3	type	type	NN
3262	84	4	system	system	NN
3262	84	5	is	be	VBZ
3262	84	6	composed	compose	VBN
3262	84	7	of	of	IN
3262	84	8	three	three	CD
3262	84	9	stages	stage	NNS
3262	84	10	:	:	:
3262	84	11	coarse	coarse	JJ
3262	84	12	-	-	HYPH
3262	84	13	level	level	NN
3262	84	14	audio	audio	JJ
3262	84	15	segmentation	segmentation	NN
3262	84	16	,	,	,
3262	84	17	fine	fine	JJ
3262	84	18	-	-	HYPH
3262	84	19	level	level	NN
3262	84	20	classification	classification	NN
3262	84	21	,	,	,
3262	84	22	and	and	CC
3262	84	23	audio	audio	JJ
3262	84	24	retrieval	retrieval	NN
3262	84	25	.	.	.
3262	85	1	In	in	IN
3262	85	2	the	the	DT
3262	85	3	first	first	JJ
3262	85	4	stage	stage	NN
3262	85	5	,	,	,
3262	85	6	audio	audio	JJ
3262	85	7	signals	signal	NNS
3262	85	8	are	be	VBP
3262	85	9	semantically	semantically	RB
3262	85	10	segmented	segmented	JJ
3262	85	11	and	and	CC
3262	85	12	classified	classify	VBD
3262	85	13	into	into	IN
3262	85	14	several	several	JJ
3262	85	15	basic	basic	JJ
3262	85	16	types	type	NNS
3262	85	17	includ-	includ-	NNP
3262	85	18	ing	ing	NNP
3262	85	19	speech	speech	NN
3262	85	20	,	,	,
3262	85	21	music	music	NN
3262	85	22	,	,	,
3262	85	23	song	song	NN
3262	85	24	,	,	,
3262	85	25	speech	speech	NN
3262	85	26	with	with	IN
3262	85	27	music	music	NN
3262	85	28	background	background	NN
3262	85	29	,	,	,
3262	85	30	environment	environment	NN
3262	85	31	sounds	sound	VBZ
3262	85	32	,	,	,
3262	85	33	and	and	CC
3262	85	34	silence	silence	NN
3262	85	35	.	.	.
3262	86	1	Some	some	DT
3262	86	2	physical	physical	JJ
3262	86	3	audio	audio	NN
3262	86	4	features	feature	NNS
3262	86	5	—	—	:
3262	86	6	such	such	JJ
3262	86	7	as	as	IN
3262	86	8	the	the	DT
3262	86	9	energy	energy	NN
3262	86	10	function	function	NN
3262	86	11	,	,	,
3262	86	12	the	the	DT
3262	86	13	fundamental	fundamental	JJ
3262	86	14	frequency	frequency	NN
3262	86	15	,	,	,
3262	86	16	and	and	CC
3262	86	17	the	the	DT
3262	86	18	spectral	spectral	JJ
3262	86	19	peak	peak	NN
3262	86	20	tracks	track	NNS
3262	86	21	—	—	:
3262	86	22	are	be	VBP
3262	86	23	examined	examine	VBN
3262	86	24	in	in	IN
3262	86	25	this	this	DT
3262	86	26	stage	stage	NN
3262	86	27	.	.	.
3262	87	1	In	in	IN
3262	87	2	the	the	DT
3262	87	3	second	second	JJ
3262	87	4	stage	stage	NN
3262	87	5	,	,	,
3262	87	6	further	further	JJ
3262	87	7	classification	classification	NN
3262	87	8	is	be	VBZ
3262	87	9	conducted	conduct	VBN
3262	87	10	for	for	IN
3262	87	11	every	every	DT
3262	87	12	basic	basic	JJ
3262	87	13	type	type	NN
3262	87	14	.	.	.
3262	88	1	Features	feature	NNS
3262	88	2	are	be	VBP
3262	88	3	extracted	extract	VBN
3262	88	4	from	from	IN
3262	88	5	the	the	DT
3262	88	6	time	time	NN
3262	88	7	-	-	HYPH
3262	88	8	frequency	frequency	NN
3262	88	9	representation	representation	NN
3262	88	10	of	of	IN
3262	88	11	audio	audio	JJ
3262	88	12	signals	signal	NNS
3262	88	13	to	to	TO
3262	88	14	reveal	reveal	VB
3262	88	15	subtle	subtle	JJ
3262	88	16	differences	difference	NNS
3262	88	17	of	of	IN
3262	88	18	timbre	timbre	NN
3262	88	19	and	and	CC
3262	88	20	pattern	pattern	NN
3262	88	21	among	among	IN
3262	88	22	different	different	JJ
3262	88	23	classes	class	NNS
3262	88	24	of	of	IN
3262	88	25	sounds	sound	NNS
3262	88	26	.	.	.
3262	89	1	Based	base	VBN
3262	89	2	on	on	IN
3262	89	3	these	these	DT
3262	89	4	differences	difference	NNS
3262	89	5	,	,	,
3262	89	6	the	the	DT
3262	89	7	coarse	coarse	JJ
3262	89	8	-	-	HYPH
3262	89	9	level	level	NN
3262	89	10	segmentation	segmentation	NN
3262	89	11	obtained	obtain	VBN
3262	89	12	in	in	IN
3262	89	13	stage	stage	NN
3262	89	14	one	one	PRP
3262	89	15	can	can	MD
3262	89	16	be	be	VB
3262	89	17	classified	classify	VBN
3262	89	18	to	to	TO
3262	89	19	narrower	narrower	VB
3262	89	20	categories	category	NNS
3262	89	21	.	.	.
3262	90	1	For	for	IN
3262	90	2	example	example	NN
3262	90	3	,	,	,
3262	90	4	speech	speech	NN
3262	90	5	can	can	MD
3262	90	6	be	be	VB
3262	90	7	differentiated	differentiate	VBN
3262	90	8	into	into	IN
3262	90	9	the	the	DT
3262	90	10	voices	voice	NNS
3262	90	11	of	of	IN
3262	90	12	men	man	NNS
3262	90	13	,	,	,
3262	90	14	women	woman	NNS
3262	90	15	,	,	,
3262	90	16	and	and	CC
3262	90	17	children	child	NNS
3262	90	18	.	.	.
3262	91	1	Finally	finally	RB
3262	91	2	,	,	,
3262	91	3	in	in	IN
3262	91	4	the	the	DT
3262	91	5	information	information	NN
3262	91	6	retrieval	retrieval	NN
3262	91	7	stage	stage	NN
3262	91	8	,	,	,
3262	91	9	two	two	CD
3262	91	10	approaches	approach	NNS
3262	91	11	—	—	:
3262	91	12	query	query	NN
3262	91	13	-	-	HYPH
3262	91	14	by	by	IN
3262	91	15	-	-	HYPH
3262	91	16	keyword	keyword	NN
3262	91	17	and	and	CC
3262	91	18	query	query	NN
3262	91	19	-	-	HYPH
3262	91	20	by	by	IN
3262	91	21	-	-	HYPH
3262	91	22	exam-	exam-	NN
3262	91	23	ple	ple	NN
3262	91	24	—	—	:
3262	91	25	are	be	VBP
3262	91	26	employed	employ	VBN
3262	91	27	.	.	.
3262	92	1	The	the	DT
3262	92	2	query	query	NN
3262	92	3	-	-	HYPH
3262	92	4	by	by	IN
3262	92	5	-	-	HYPH
3262	92	6	keyword	keyword	NN
3262	92	7	approach	approach	NN
3262	92	8	is	be	VBZ
3262	92	9	more	more	JJR
3262	92	10	like	like	IN
3262	92	11	the	the	DT
3262	92	12	traditional	traditional	JJ
3262	92	13	text	text	NN
3262	92	14	-	-	HYPH
3262	92	15	based	base	VBN
3262	92	16	search	search	NN
3262	92	17	system	system	NN
3262	92	18	.	.	.
3262	93	1	The	the	DT
3262	93	2	query	query	NN
3262	93	3	-	-	HYPH
3262	93	4	by	by	IN
3262	93	5	-	-	HYPH
3262	93	6	example	example	NN
3262	93	7	approach	approach	NN
3262	93	8	is	be	VBZ
3262	93	9	similar	similar	JJ
3262	93	10	to	to	IN
3262	93	11	content	content	NN
3262	93	12	-	-	HYPH
3262	93	13	based	base	VBN
3262	93	14	image	image	NN
3262	93	15	retrieval	retrieval	NN
3262	93	16	systems	system	NNS
3262	93	17	where	where	WRB
3262	93	18	an	an	DT
3262	93	19	image	image	NN
3262	93	20	can	can	MD
3262	93	21	be	be	VB
3262	93	22	searched	search	VBN
3262	93	23	by	by	IN
3262	93	24	color	color	NN
3262	93	25	,	,	,
3262	93	26	texture	texture	NN
3262	93	27	,	,	,
3262	93	28	and	and	CC
3262	93	29	histogram	histogram	NNP
3262	93	30	,	,	,
3262	93	31	and	and	CC
3262	93	32	audio	audio	JJ
3262	93	33	clips	clip	NNS
3262	93	34	can	can	MD
3262	93	35	be	be	VB
3262	93	36	retrieved	retrieve	VBN
3262	93	37	with	with	IN
3262	93	38	distinct	distinct	JJ
3262	93	39	features	feature	NNS
3262	93	40	,	,	,
3262	93	41	such	such	JJ
3262	93	42	as	as	IN
3262	93	43	timbre	timbre	NNP
3262	93	44	,	,	,
3262	93	45	pitch	pitch	NN
3262	93	46	,	,	,
3262	93	47	and	and	CC
3262	93	48	rhythm	rhythm	NN
3262	93	49	.	.	.
3262	94	1	This	this	DT
3262	94	2	way	way	NN
3262	94	3	,	,	,
3262	94	4	a	a	DT
3262	94	5	user	user	NN
3262	94	6	may	may	MD
3262	94	7	choose	choose	VB
3262	94	8	from	from	IN
3262	94	9	a	a	DT
3262	94	10	given	give	VBN
3262	94	11	list	list	NN
3262	94	12	of	of	IN
3262	94	13	features	feature	NNS
3262	94	14	,	,	,
3262	94	15	listen	listen	VB
3262	94	16	to	to	IN
3262	94	17	the	the	DT
3262	94	18	retrieved	retrieve	VBN
3262	94	19	samples	sample	NNS
3262	94	20	,	,	,
3262	94	21	and	and	CC
3262	94	22	modify	modify	VB
3262	94	23	the	the	DT
3262	94	24	input	input	NN
3262	94	25	feature	feature	NN
3262	94	26	set	set	VBN
3262	94	27	to	to	TO
3262	94	28	get	get	VB
3262	94	29	more	more	RBR
3262	94	30	desired	desire	VBN
3262	94	31	results	result	NNS
3262	94	32	.	.	.
3262	95	1	Zhang	Zhang	NNP
3262	95	2	and	and	CC
3262	95	3	Kuo	Kuo	NNP
3262	95	4	’s	’s	POS
3262	95	5	prototype	prototype	NN
3262	95	6	is	be	VBZ
3262	95	7	a	a	DT
3262	95	8	very	very	RB
3262	95	9	typical	typical	JJ
3262	95	10	and	and	CC
3262	95	11	classic	classic	JJ
3262	95	12	CBAR	CBAR	NNP
3262	95	13	system	system	NN
3262	95	14	.	.	.
3262	96	1	It	-PRON-	PRP
3262	96	2	is	be	VBZ
3262	96	3	relatively	relatively	RB
3262	96	4	mature	mature	JJ
3262	96	5	and	and	CC
3262	96	6	can	can	MD
3262	96	7	be	be	VB
3262	96	8	used	use	VBN
3262	96	9	by	by	IN
3262	96	10	large	large	JJ
3262	96	11	digital	digital	JJ
3262	96	12	audio	audio	JJ
3262	96	13	libraries	library	NNS
3262	96	14	.	.	.
3262	97	1	More	more	RBR
3262	97	2	recently	recently	RB
3262	97	3	,	,	,
3262	97	4	Li	Li	NNP
3262	97	5	et	et	NNP
3262	97	6	al	al	NNP
3262	97	7	.	.	.
3262	98	1	(	(	-LRB-
3262	98	2	2003	2003	CD
3262	98	3	)	)	-RRB-
3262	98	4	proposed	propose	VBD
3262	98	5	a	a	DT
3262	98	6	new	new	JJ
3262	98	7	feature	feature	NN
3262	98	8	extraction	extraction	NN
3262	98	9	method	method	NN
3262	98	10	particularly	particularly	RB
3262	98	11	for	for	IN
3262	98	12	music	music	NN
3262	98	13	genre	genre	NN
3262	98	14	classifica-	classifica-	NN
3262	98	15	tion	tion	NN
3262	98	16	named	name	VBN
3262	98	17	Daubechies	Daubechies	NNP
3262	98	18	Wavelet	Wavelet	NNP
3262	98	19	Coefficient	Coefficient	NNP
3262	98	20	Histograms	Histograms	NNPS
3262	98	21	(	(	-LRB-
3262	98	22	DWCHs	DWCHs	NNP
3262	98	23	)	)	-RRB-
3262	98	24	.	.	.
3262	99	1	DWCHs	dwch	NNS
3262	99	2	capture	capture	VBP
3262	99	3	the	the	DT
3262	99	4	local	local	JJ
3262	99	5	and	and	CC
3262	99	6	global	global	JJ
3262	99	7	informa-	informa-	NN
3262	99	8	tion	tion	NN
3262	99	9	of	of	IN
3262	99	10	music	music	NN
3262	99	11	signals	signal	NNS
3262	99	12	simultaneously	simultaneously	RB
3262	99	13	by	by	IN
3262	99	14	computing	compute	VBG
3262	99	15	their	-PRON-	PRP$
3262	99	16	histograms	histogram	NNS
3262	99	17	.	.	.
3262	100	1	Similar	similar	JJ
3262	100	2	to	to	IN
3262	100	3	other	other	JJ
3262	100	4	CBAR	CBAR	NNP
3262	100	5	strategies	strategy	NNS
3262	100	6	,	,	,
3262	100	7	this	this	DT
3262	100	8	method	method	NN
3262	100	9	divides	divide	VBZ
3262	100	10	the	the	DT
3262	100	11	process	process	NN
3262	100	12	of	of	IN
3262	100	13	music	music	NN
3262	100	14	genre	genre	NN
3262	100	15	classification	classification	NN
3262	100	16	into	into	IN
3262	100	17	two	two	CD
3262	100	18	steps	step	NNS
3262	100	19	:	:	:
3262	100	20	feature	feature	NN
3262	100	21	extraction	extraction	NN
3262	100	22	and	and	CC
3262	100	23	multi	multi	JJ
3262	100	24	-	-	JJ
3262	100	25	class	class	JJ
3262	100	26	classification	classification	NN
3262	100	27	.	.	.
3262	101	1	The	the	DT
3262	101	2	music	music	NN
3262	101	3	signal	signal	NN
3262	101	4	information	information	NN
3262	101	5	representing	represent	VBG
3262	101	6	the	the	DT
3262	101	7	music	music	NN
3262	101	8	is	be	VBZ
3262	101	9	extracted	extract	VBN
3262	101	10	first	first	RB
3262	101	11	,	,	,
3262	101	12	and	and	CC
3262	101	13	then	then	RB
3262	101	14	an	an	DT
3262	101	15	algorithm	algorithm	NN
3262	101	16	is	be	VBZ
3262	101	17	used	use	VBN
3262	101	18	to	to	TO
3262	101	19	identify	identify	VB
3262	101	20	the	the	DT
3262	101	21	labels	label	NNS
3262	101	22	from	from	IN
3262	101	23	the	the	DT
3262	101	24	representation	representation	NN
3262	101	25	of	of	IN
3262	101	26	the	the	DT
3262	101	27	music	music	NN
3262	101	28	sounds	sound	VBZ
3262	101	29	with	with	IN
3262	101	30	respect	respect	NN
3262	101	31	to	to	IN
3262	101	32	their	-PRON-	PRP$
3262	101	33	features	feature	NNS
3262	101	34	.	.	.
3262	102	1	Since	since	IN
3262	102	2	the	the	DT
3262	102	3	decomposition	decomposition	NN
3262	102	4	of	of	IN
3262	102	5	audio	audio	JJ
3262	102	6	signal	signal	NN
3262	102	7	can	can	MD
3262	102	8	produce	produce	VB
3262	102	9	a	a	DT
3262	102	10	set	set	NN
3262	102	11	of	of	IN
3262	102	12	subband	subband	JJ
3262	102	13	signals	signal	NNS
3262	102	14	at	at	IN
3262	102	15	different	different	JJ
3262	102	16	frequencies	frequency	NNS
3262	102	17	cor-	cor-	IN
3262	102	18	responding	respond	VBG
3262	102	19	to	to	IN
3262	102	20	different	different	JJ
3262	102	21	characteristics	characteristic	NNS
3262	102	22	,	,	,
3262	102	23	Li	Li	NNP
3262	102	24	et	et	FW
3262	102	25	al	al	NNP
3262	102	26	.	.	.
3262	103	1	(	(	-LRB-
3262	103	2	2003	2003	CD
3262	103	3	)	)	-RRB-
3262	103	4	proposed	propose	VBD
3262	103	5	a	a	DT
3262	103	6	new	new	JJ
3262	103	7	methodology	methodology	NN
3262	103	8	,	,	,
3262	103	9	the	the	DT
3262	103	10	DWCHs	DWCHs	NNP
3262	103	11	algorithm	algorithm	NN
3262	103	12	,	,	,
3262	103	13	for	for	IN
3262	103	14	feature	feature	NN
3262	103	15	extraction	extraction	NN
3262	103	16	.	.	.
3262	104	1	With	with	IN
3262	104	2	this	this	DT
3262	104	3	algorithm	algorithm	NN
3262	104	4	,	,	,
3262	104	5	the	the	DT
3262	104	6	decomposi-	decomposi-	NNP
3262	104	7	tion	tion	NN
3262	104	8	of	of	IN
3262	104	9	the	the	DT
3262	104	10	music	music	NN
3262	104	11	signals	signal	NNS
3262	104	12	is	be	VBZ
3262	104	13	obtained	obtain	VBN
3262	104	14	at	at	IN
3262	104	15	the	the	DT
3262	104	16	beginning	beginning	NN
3262	104	17	,	,	,
3262	104	18	and	and	CC
3262	104	19	then	then	RB
3262	104	20	a	a	DT
3262	104	21	histogram	histogram	NN
3262	104	22	of	of	IN
3262	104	23	each	each	DT
3262	104	24	subband	subband	NN
3262	104	25	is	be	VBZ
3262	104	26	constructed	construct	VBN
3262	104	27	.	.	.
3262	105	1	Hence	hence	RB
3262	105	2	,	,	,
3262	105	3	the	the	DT
3262	105	4	energy	energy	NN
3262	105	5	for	for	IN
3262	105	6	each	each	DT
3262	105	7	subband	subband	NN
3262	105	8	is	be	VBZ
3262	105	9	computed	compute	VBN
3262	105	10	,	,	,
3262	105	11	and	and	CC
3262	105	12	the	the	DT
3262	105	13	charac-	charac-	JJ
3262	105	14	teristics	teristic	NNS
3262	105	15	of	of	IN
3262	105	16	the	the	DT
3262	105	17	music	music	NN
3262	105	18	are	be	VBP
3262	105	19	represented	represent	VBN
3262	105	20	by	by	IN
3262	105	21	these	these	DT
3262	105	22	subbands	subband	NNS
3262	105	23	.	.	.
3262	106	1	One	one	CD
3262	106	2	finding	find	VBG
3262	106	3	from	from	IN
3262	106	4	this	this	DT
3262	106	5	research	research	NN
3262	106	6	reveals	reveal	VBZ
3262	106	7	that	that	IN
3262	106	8	this	this	DT
3262	106	9	methodol-	methodol-	JJ
3262	106	10	ogy	ogy	NN
3262	106	11	,	,	,
3262	106	12	along	along	IN
3262	106	13	with	with	IN
3262	106	14	advanced	advanced	JJ
3262	106	15	machine	machine	NN
3262	106	16	learning	learning	NN
3262	106	17	techniques	technique	NNS
3262	106	18	,	,	,
3262	106	19	has	have	VBZ
3262	106	20	significantly	significantly	RB
3262	106	21	improved	improve	VBN
3262	106	22	accuracy	accuracy	NN
3262	106	23	of	of	IN
3262	106	24	music	music	NN
3262	106	25	genre	genre	NN
3262	106	26	clas-	clas-	JJ
3262	106	27	sification	sification	NN
3262	106	28	(	(	-LRB-
3262	106	29	Li	Li	NNP
3262	106	30	et	et	FW
3262	106	31	al	al	NNP
3262	106	32	.	.	.
3262	107	1	2003	2003	CD
3262	107	2	)	)	-RRB-
3262	107	3	.	.	.
3262	108	1	Therefore	therefore	RB
3262	108	2	,	,	,
3262	108	3	this	this	DT
3262	108	4	methodology	methodology	NN
3262	108	5	potentially	potentially	RB
3262	108	6	can	can	MD
3262	108	7	be	be	VB
3262	108	8	used	use	VBN
3262	108	9	by	by	IN
3262	108	10	those	those	DT
3262	108	11	digital	digital	JJ
3262	108	12	music	music	NN
3262	108	13	libraries	library	NNS
3262	108	14	widely	widely	RB
3262	108	15	developed	develop	VBN
3262	108	16	in	in	IN
3262	108	17	past	past	JJ
3262	108	18	several	several	JJ
3262	108	19	years	year	NNS
3262	108	20	.	.	.
3262	109	1	■	■	NFP
3262	109	2	Content	content	NN
3262	109	3	-	-	HYPH
3262	109	4	based	base	VBN
3262	109	5	video	video	NN
3262	109	6	retrieval	retrieval	NN
3262	109	7	Content	Content	NNP
3262	109	8	-	-	HYPH
3262	109	9	based	base	VBN
3262	109	10	video	video	NN
3262	109	11	retrieval	retrieval	NN
3262	109	12	(	(	-LRB-
3262	109	13	CBVR	CBVR	NNP
3262	109	14	)	)	-RRB-
3262	109	15	is	be	VBZ
3262	109	16	a	a	DT
3262	109	17	more	more	RBR
3262	109	18	recent	recent	JJ
3262	109	19	research	research	NN
3262	109	20	topic	topic	NN
3262	109	21	than	than	IN
3262	109	22	CBIR	CBIR	NNP
3262	109	23	and	and	CC
3262	109	24	CBAR	CBAR	NNP
3262	109	25	,	,	,
3262	109	26	partly	partly	RB
3262	109	27	because	because	IN
3262	109	28	the	the	DT
3262	109	29	digitization	digitization	NN
3262	109	30	technology	technology	NN
3262	109	31	for	for	IN
3262	109	32	video	video	NN
3262	109	33	appeared	appear	VBD
3262	109	34	later	later	RB
3262	109	35	than	than	IN
3262	109	36	those	those	DT
3262	109	37	for	for	IN
3262	109	38	image	image	NN
3262	109	39	and	and	CC
3262	109	40	audio	audio	NN
3262	109	41	.	.	.
3262	110	1	As	as	IN
3262	110	2	digital	digital	JJ
3262	110	3	video	video	NN
3262	110	4	Websites	website	NNS
3262	110	5	such	such	JJ
3262	110	6	as	as	IN
3262	110	7	YouTube	YouTube	NNP
3262	110	8	and	and	CC
3262	110	9	Google	Google	NNP
3262	110	10	Video	Video	NNP
3262	110	11	become	become	VBP
3262	110	12	more	more	RBR
3262	110	13	popular	popular	JJ
3262	110	14	,	,	,
3262	110	15	how	how	WRB
3262	110	16	to	to	TO
3262	110	17	retrieve	retrieve	VB
3262	110	18	desired	desire	VBN
3262	110	19	video	video	NN
3262	110	20	clips	clip	NNS
3262	110	21	effectively	effectively	RB
3262	110	22	is	be	VBZ
3262	110	23	a	a	DT
3262	110	24	great	great	JJ
3262	110	25	con-	con-	NN
3262	110	26	cern	cern	NN
3262	110	27	.	.	.
3262	111	1	Searching	search	VBG
3262	111	2	by	by	IN
3262	111	3	some	some	DT
3262	111	4	features	feature	NNS
3262	111	5	of	of	IN
3262	111	6	video	video	NN
3262	111	7	,	,	,
3262	111	8	such	such	JJ
3262	111	9	as	as	IN
3262	111	10	motion	motion	NN
3262	111	11	and	and	CC
3262	111	12	texture	texture	NN
3262	111	13	,	,	,
3262	111	14	can	can	MD
3262	111	15	be	be	VB
3262	111	16	a	a	DT
3262	111	17	good	good	JJ
3262	111	18	complement	complement	NN
3262	111	19	to	to	IN
3262	111	20	the	the	DT
3262	111	21	traditional	traditional	JJ
3262	111	22	text	text	NN
3262	111	23	-	-	HYPH
3262	111	24	based	base	VBN
3262	111	25	search	search	NN
3262	111	26	method	method	NN
3262	111	27	.	.	.
3262	112	1	One	one	CD
3262	112	2	of	of	IN
3262	112	3	the	the	DT
3262	112	4	earliest	early	JJS
3262	112	5	examples	example	NNS
3262	112	6	is	be	VBZ
3262	112	7	the	the	DT
3262	112	8	VideoQ	videoq	JJ
3262	112	9	system	system	NN
3262	112	10	developed	develop	VBN
3262	112	11	by	by	IN
3262	112	12	Chang	Chang	NNP
3262	112	13	et	et	FW
3262	112	14	al	al	NNP
3262	112	15	.	.	.
3262	113	1	(	(	-LRB-
3262	113	2	1997	1997	CD
3262	113	3	)	)	-RRB-
3262	113	4	of	of	IN
3262	113	5	Columbia	Columbia	NNP
3262	113	6	University	University	NNP
3262	113	7	(	(	-LRB-
3262	113	8	www.ctr.columbia.edu/VideoQ	www.ctr.columbia.edu/VideoQ	NNP
3262	113	9	)	)	-RRB-
3262	113	10	,	,	,
3262	113	11	which	which	WDT
3262	113	12	allows	allow	VBZ
3262	113	13	a	a	DT
3262	113	14	user	user	NN
3262	113	15	to	to	TO
3262	113	16	search	search	VB
3262	113	17	video	video	NN
3262	113	18	based	base	VBN
3262	113	19	on	on	IN
3262	113	20	a	a	DT
3262	113	21	rich	rich	JJ
3262	113	22	set	set	NN
3262	113	23	of	of	IN
3262	113	24	visual	visual	JJ
3262	113	25	features	feature	NNS
3262	113	26	and	and	CC
3262	113	27	spatio	spatio	NN
3262	113	28	-	-	HYPH
3262	113	29	temporal	temporal	JJ
3262	113	30	relationships	relationship	NNS
3262	113	31	.	.	.
3262	114	1	The	the	DT
3262	114	2	video	video	NN
3262	114	3	clips	clip	NNS
3262	114	4	in	in	IN
3262	114	5	the	the	DT
3262	114	6	data-	data-	JJ
3262	114	7	base	base	NN
3262	114	8	are	be	VBP
3262	114	9	stored	store	VBN
3262	114	10	as	as	IN
3262	114	11	MPEG	mpeg	NN
3262	114	12	files	file	NNS
3262	114	13	.	.	.
3262	115	1	Through	through	IN
3262	115	2	a	a	DT
3262	115	3	Web	web	NN
3262	115	4	interface	interface	NN
3262	115	5	,	,	,
3262	115	6	the	the	DT
3262	115	7	user	user	NN
3262	115	8	can	can	MD
3262	115	9	formulate	formulate	VB
3262	115	10	a	a	DT
3262	115	11	query	query	NN
3262	115	12	scene	scene	NN
3262	115	13	as	as	IN
3262	115	14	a	a	DT
3262	115	15	collection	collection	NN
3262	115	16	of	of	IN
3262	115	17	objects	object	NNS
3262	115	18	with	with	IN
3262	115	19	different	different	JJ
3262	115	20	attributes	attribute	NNS
3262	115	21	,	,	,
3262	115	22	including	include	VBG
3262	115	23	motion	motion	NN
3262	115	24	,	,	,
3262	115	25	shape	shape	NN
3262	115	26	,	,	,
3262	115	27	color	color	NN
3262	115	28	,	,	,
3262	115	29	and	and	CC
3262	115	30	texture	texture	NN
3262	115	31	.	.	.
3262	116	1	Once	once	IN
3262	116	2	the	the	DT
3262	116	3	user	user	NN
3262	116	4	has	have	VBZ
3262	116	5	formulated	formulate	VBN
3262	116	6	the	the	DT
3262	116	7	query	query	NN
3262	116	8	,	,	,
3262	116	9	it	-PRON-	PRP
3262	116	10	is	be	VBZ
3262	116	11	sent	send	VBN
3262	116	12	to	to	IN
3262	116	13	a	a	DT
3262	116	14	query	query	NN
3262	116	15	server	server	NN
3262	116	16	,	,	,
3262	116	17	which	which	WDT
3262	116	18	contains	contain	VBZ
3262	116	19	several	several	JJ
3262	116	20	databases	database	NNS
3262	116	21	for	for	IN
3262	116	22	different	different	JJ
3262	116	23	content	content	NN
3262	116	24	features	feature	NNS
3262	116	25	.	.	.
3262	117	1	On	on	IN
3262	117	2	the	the	DT
3262	117	3	query	query	NN
3262	117	4	server	server	NN
3262	117	5	,	,	,
3262	117	6	the	the	DT
3262	117	7	similarities	similarity	NNS
3262	117	8	between	between	IN
3262	117	9	the	the	DT
3262	117	10	features	feature	NNS
3262	117	11	of	of	IN
3262	117	12	each	each	DT
3262	117	13	object	object	NN
3262	117	14	specified	specify	VBN
3262	117	15	in	in	IN
3262	117	16	the	the	DT
3262	117	17	query	query	NN
3262	117	18	and	and	CC
3262	117	19	those	those	DT
3262	117	20	of	of	IN
3262	117	21	the	the	DT
3262	117	22	objects	object	NNS
3262	117	23	in	in	IN
3262	117	24	the	the	DT
3262	117	25	database	database	NN
3262	117	26	are	be	VBP
3262	117	27	com-	com-	RB
3262	117	28	puted	put	VBN
3262	117	29	;	;	:
3262	117	30	a	a	DT
3262	117	31	list	list	NN
3262	117	32	of	of	IN
3262	117	33	video	video	NN
3262	117	34	clips	clip	NNS
3262	117	35	is	be	VBZ
3262	117	36	then	then	RB
3262	117	37	retrieved	retrieve	VBN
3262	117	38	based	base	VBN
3262	117	39	on	on	IN
3262	117	40	their	-PRON-	PRP$
3262	117	41	similarity	similarity	NN
3262	117	42	values	value	NNS
3262	117	43	.	.	.
3262	118	1	For	for	IN
3262	118	2	each	each	DT
3262	118	3	of	of	IN
3262	118	4	these	these	DT
3262	118	5	video	video	NN
3262	118	6	clips	clip	NNS
3262	118	7	,	,	,
3262	118	8	key	key	NN
3262	118	9	-	-	HYPH
3262	118	10	frames	frame	NNS
3262	118	11	are	be	VBP
3262	118	12	dynamically	dynamically	RB
3262	118	13	extracted	extract	VBN
3262	118	14	from	from	IN
3262	118	15	the	the	DT
3262	118	16	video	video	NN
3262	118	17	database	database	NN
3262	118	18	and	and	CC
3262	118	19	returned	return	VBD
3262	118	20	to	to	IN
3262	118	21	browser	browser	NN
3262	118	22	.	.	.
3262	119	1	The	the	DT
3262	119	2	matched	match	VBN
3262	119	3	objects	object	NNS
3262	119	4	are	be	VBP
3262	119	5	highlighted	highlight	VBN
3262	119	6	in	in	IN
3262	119	7	the	the	DT
3262	119	8	returned	return	VBN
3262	119	9	key	key	NN
3262	119	10	-	-	HYPH
3262	119	11	frame	frame	NN
3262	119	12	.	.	.
3262	120	1	The	the	DT
3262	120	2	user	user	NN
3262	120	3	can	can	MD
3262	120	4	interactively	interactively	RB
3262	120	5	view	view	VB
3262	120	6	these	these	DT
3262	120	7	matched	match	VBN
3262	120	8	video	video	NN
3262	120	9	clips	clip	NNS
3262	120	10	by	by	IN
3262	120	11	simply	simply	RB
3262	120	12	clicking	click	VBG
3262	120	13	on	on	IN
3262	120	14	the	the	DT
3262	120	15	key-	key-	NN
3262	120	16	frame	frame	NN
3262	120	17	.	.	.
3262	121	1	Meanwhile	meanwhile	RB
3262	121	2	,	,	,
3262	121	3	the	the	DT
3262	121	4	video	video	NN
3262	121	5	clip	clip	NN
3262	121	6	corresponding	correspond	VBG
3262	121	7	to	to	IN
3262	121	8	that	that	DT
3262	121	9	key	key	JJ
3262	121	10	-	-	HYPH
3262	121	11	frame	frame	NN
3262	121	12	is	be	VBZ
3262	121	13	extracted	extract	VBN
3262	121	14	from	from	IN
3262	121	15	the	the	DT
3262	121	16	video	video	NN
3262	121	17	database	database	NN
3262	121	18	(	(	-LRB-
3262	121	19	Chang	Chang	NNP
3262	121	20	et	et	FW
3262	121	21	al	al	NNP
3262	121	22	.	.	.
3262	122	1	1997	1997	CD
3262	122	2	)	)	-RRB-
3262	122	3	.	.	.
3262	123	1	Figures	figure	NNS
3262	123	2	5.1–5.2	5.1–5.2	CD
3262	123	3	show	show	VBP
3262	123	4	an	an	DT
3262	123	5	example	example	NN
3262	123	6	of	of	IN
3262	123	7	a	a	DT
3262	123	8	visual	visual	JJ
3262	123	9	search	search	NN
3262	123	10	through	through	IN
3262	123	11	the	the	DT
3262	123	12	VideoQ	videoq	JJ
3262	123	13	system	system	NN
3262	123	14	.	.	.
3262	124	1	Many	many	JJ
3262	124	2	other	other	JJ
3262	124	3	CBVR	CBVR	NNP
3262	124	4	projects	project	NNS
3262	124	5	also	also	RB
3262	124	6	examine	examine	VBP
3262	124	7	these	these	DT
3262	124	8	content	content	NN
3262	124	9	features	feature	NNS
3262	124	10	and	and	CC
3262	124	11	try	try	VBP
3262	124	12	to	to	TO
3262	124	13	find	find	VB
3262	124	14	more	more	RBR
3262	124	15	efficient	efficient	JJ
3262	124	16	ways	way	NNS
3262	124	17	to	to	TO
3262	124	18	retrieve	retrieve	VB
3262	124	19	data	datum	NNS
3262	124	20	.	.	.
3262	125	1	A	a	DT
3262	125	2	recent	recent	JJ
3262	125	3	example	example	NN
3262	125	4	is	be	VBZ
3262	125	5	Wang	Wang	NNP
3262	125	6	et	et	NNP
3262	125	7	al	al	NNP
3262	125	8	.	.	.
3262	125	9	’s	’s	NNP
3262	125	10	(	(	-LRB-
3262	125	11	2006	2006	CD
3262	125	12	)	)	-RRB-
3262	125	13	project	project	NN
3262	125	14	,	,	,
3262	125	15	Vferret	Vferret	NNP
3262	125	16	,	,	,
3262	125	17	a	a	DT
3262	125	18	content	content	NN
3262	125	19	-	-	HYPH
3262	125	20	based	base	VBN
3262	125	21	similarity	similarity	NN
3262	125	22	search	search	NN
3262	125	23	tool	tool	NN
3262	125	24	for	for	IN
3262	125	25	continu-	continu-	NNP
3262	125	26	ous	ous	NNP
3262	125	27	archived	archived	JJ
3262	125	28	video	video	NN
3262	125	29	.	.	.
3262	126	1	The	the	DT
3262	126	2	Vferret	Vferret	NNP
3262	126	3	system	system	NN
3262	126	4	segments	segment	VBZ
3262	126	5	video	video	NN
3262	126	6	data	datum	NNS
3262	126	7	into	into	IN
3262	126	8	clips	clip	NNS
3262	126	9	and	and	CC
3262	126	10	extracts	extract	VBZ
3262	126	11	both	both	CC
3262	126	12	visual	visual	JJ
3262	126	13	and	and	CC
3262	126	14	audio	audio	JJ
3262	126	15	features	feature	NNS
3262	126	16	as	as	IN
3262	126	17	metadata	metadata	NN
3262	126	18	.	.	.
3262	127	1	Then	then	RB
3262	127	2	a	a	DT
3262	127	3	user	user	NN
3262	127	4	can	can	MD
3262	127	5	do	do	VB
3262	127	6	a	a	DT
3262	127	7	metadata	metadata	NN
3262	127	8	search	search	NN
3262	127	9	or	or	CC
3262	127	10	aRtiCLE	article	CD
3262	127	11	titLE	title	CD
3262	127	12	|	|	NNP
3262	127	13	autHoR	autHoR	NNP
3262	127	14	45ContEnt	45ContEnt	NNP
3262	127	15	-	-	HYPH
3262	127	16	BasED	BasED	NNP
3262	127	17	inFoRmation	information	NN
3262	127	18	REtRiEvaL	REtRiEvaL	NNP
3262	127	19	anD	and	CC
3262	127	20	DiGitaL	DiGitaL	NNP
3262	127	21	LiBRaRiEs	LiBRaRiEs	NNP
3262	127	22	|	|	NNP
3262	127	23	Wan	Wan	NNP
3262	127	24	anD	and	CC
3262	127	25	Liu	Liu	NNP
3262	127	26	45	45	CD
3262	127	27	content	content	NN
3262	127	28	-	-	HYPH
3262	127	29	based	base	VBN
3262	127	30	search	search	NN
3262	127	31	to	to	TO
3262	127	32	retrieve	retrieve	VB
3262	127	33	desired	desire	VBN
3262	127	34	video	video	NN
3262	127	35	clips	clip	NNS
3262	127	36	.	.	.
3262	128	1	In	in	IN
3262	128	2	the	the	DT
3262	128	3	first	first	JJ
3262	128	4	stage	stage	NN
3262	128	5	,	,	,
3262	128	6	a	a	DT
3262	128	7	simple	simple	JJ
3262	128	8	segmentation	segmentation	NN
3262	128	9	method	method	NN
3262	128	10	is	be	VBZ
3262	128	11	used	use	VBN
3262	128	12	to	to	TO
3262	128	13	split	split	VB
3262	128	14	the	the	DT
3262	128	15	archived	archived	JJ
3262	128	16	digital	digital	JJ
3262	128	17	video	video	NN
3262	128	18	into	into	IN
3262	128	19	five	five	CD
3262	128	20	-	-	HYPH
3262	128	21	minute	minute	NN
3262	128	22	video	video	NN
3262	128	23	clips	clip	NNS
3262	128	24	.	.	.
3262	129	1	The	the	DT
3262	129	2	system	system	NN
3262	129	3	then	then	RB
3262	129	4	extracts	extract	VBZ
3262	129	5	twenty	twenty	CD
3262	129	6	image	image	NN
3262	129	7	frames	frame	NNS
3262	129	8	evenly	evenly	RB
3262	129	9	from	from	IN
3262	129	10	each	each	DT
3262	129	11	of	of	IN
3262	129	12	these	these	DT
3262	129	13	five	five	CD
3262	129	14	-	-	HYPH
3262	129	15	minute	minute	NN
3262	129	16	video	video	NN
3262	129	17	clips	clip	NNS
3262	129	18	for	for	IN
3262	129	19	visual	visual	JJ
3262	129	20	feature	feature	NN
3262	129	21	extraction	extraction	NN
3262	129	22	.	.	.
3262	130	1	Additionally	additionally	RB
3262	130	2	,	,	,
3262	130	3	the	the	DT
3262	130	4	system	system	NN
3262	130	5	splits	split	VBZ
3262	130	6	the	the	DT
3262	130	7	audio	audio	JJ
3262	130	8	channel	channel	NN
3262	130	9	of	of	IN
3262	130	10	each	each	DT
3262	130	11	clip	clip	NN
3262	130	12	into	into	IN
3262	130	13	twenty	twenty	CD
3262	130	14	individual	individual	JJ
3262	130	15	fifteen-	fifteen-	NN
3262	130	16	second	second	JJ
3262	130	17	segments	segment	NNS
3262	130	18	for	for	IN
3262	130	19	further	further	JJ
3262	130	20	audio	audio	JJ
3262	130	21	feature	feature	NN
3262	130	22	extraction	extraction	NN
3262	130	23	.	.	.
3262	131	1	In	in	IN
3262	131	2	the	the	DT
3262	131	3	second	second	JJ
3262	131	4	stage	stage	NN
3262	131	5	,	,	,
3262	131	6	both	both	CC
3262	131	7	audio	audio	JJ
3262	131	8	and	and	CC
3262	131	9	visual	visual	JJ
3262	131	10	features	feature	NNS
3262	131	11	are	be	VBP
3262	131	12	extracted	extract	VBN
3262	131	13	.	.	.
3262	132	1	For	for	IN
3262	132	2	visual	visual	JJ
3262	132	3	features	feature	NNS
3262	132	4	,	,	,
3262	132	5	the	the	DT
3262	132	6	color	color	NN
3262	132	7	element	element	NN
3262	132	8	is	be	VBZ
3262	132	9	used	use	VBN
3262	132	10	as	as	IN
3262	132	11	the	the	DT
3262	132	12	content	content	NN
3262	132	13	feature	feature	NN
3262	132	14	.	.	.
3262	133	1	For	for	IN
3262	133	2	audio	audio	JJ
3262	133	3	features	feature	NNS
3262	133	4	,	,	,
3262	133	5	154	154	CD
3262	133	6	audio	audio	NN
3262	133	7	fea-	fea-	:
3262	133	8	tures	ture	NNS
3262	133	9	originally	originally	RB
3262	133	10	used	use	VBN
3262	133	11	by	by	IN
3262	133	12	Ellis	Ellis	NNP
3262	133	13	and	and	CC
3262	133	14	Lee	Lee	NNP
3262	133	15	(	(	-LRB-
3262	133	16	2004	2004	CD
3262	133	17	)	)	-RRB-
3262	133	18	to	to	TO
3262	133	19	describe	describe	VB
3262	133	20	audio	audio	JJ
3262	133	21	segments	segment	NNS
3262	133	22	are	be	VBP
3262	133	23	computed	compute	VBN
3262	133	24	.	.	.
3262	134	1	For	for	IN
3262	134	2	each	each	DT
3262	134	3	fifteen	fifteen	CD
3262	134	4	-	-	HYPH
3262	134	5	second	second	JJ
3262	134	6	video	video	NN
3262	134	7	segment	segment	NN
3262	134	8	,	,	,
3262	134	9	the	the	DT
3262	134	10	visual	visual	JJ
3262	134	11	feature	feature	NN
3262	134	12	vector	vector	NN
3262	134	13	extracted	extract	VBN
3262	134	14	from	from	IN
3262	134	15	the	the	DT
3262	134	16	sample	sample	NN
3262	134	17	image	image	NN
3262	134	18	and	and	CC
3262	134	19	the	the	DT
3262	134	20	audio	audio	JJ
3262	134	21	feature	feature	NN
3262	134	22	vector	vector	NN
3262	134	23	extracted	extract	VBN
3262	134	24	from	from	IN
3262	134	25	the	the	DT
3262	134	26	corresponding	corresponding	JJ
3262	134	27	audio	audio	NN
3262	134	28	segment	segment	NN
3262	134	29	are	be	VBP
3262	134	30	combined	combine	VBN
3262	134	31	into	into	IN
3262	134	32	a	a	DT
3262	134	33	single	single	JJ
3262	134	34	feature	feature	NN
3262	134	35	vector	vector	NN
3262	134	36	.	.	.
3262	135	1	In	in	IN
3262	135	2	the	the	DT
3262	135	3	information	information	NN
3262	135	4	retrieval	retrieval	NN
3262	135	5	stage	stage	NN
3262	135	6	,	,	,
3262	135	7	the	the	DT
3262	135	8	user	user	NN
3262	135	9	submits	submit	VBZ
3262	135	10	a	a	DT
3262	135	11	video	video	NN
3262	135	12	clip	clip	NN
3262	135	13	query	query	NN
3262	135	14	at	at	IN
3262	135	15	first	first	RB
3262	135	16	,	,	,
3262	135	17	then	then	RB
3262	135	18	its	-PRON-	PRP$
3262	135	19	feature	feature	NN
3262	135	20	vector	vector	NN
3262	135	21	is	be	VBZ
3262	135	22	computed	compute	VBN
3262	135	23	and	and	CC
3262	135	24	compared	compare	VBN
3262	135	25	with	with	IN
3262	135	26	that	that	DT
3262	135	27	of	of	IN
3262	135	28	video	video	NN
3262	135	29	clips	clip	NNS
3262	135	30	in	in	IN
3262	135	31	the	the	DT
3262	135	32	database	database	NN
3262	135	33	,	,	,
3262	135	34	and	and	CC
3262	135	35	the	the	DT
3262	135	36	most	most	RBS
3262	135	37	similar	similar	JJ
3262	135	38	clips	clip	NNS
3262	135	39	are	be	VBP
3262	135	40	retrieved	retrieve	VBN
3262	135	41	(	(	-LRB-
3262	135	42	Wang	Wang	NNP
3262	135	43	et	et	NNP
3262	135	44	al	al	NNP
3262	135	45	.	.	.
3262	136	1	2006	2006	CD
3262	136	2	)	)	-RRB-
3262	136	3	.	.	.
3262	137	1	Similar	similar	JJ
3262	137	2	projects	project	NNS
3262	137	3	in	in	IN
3262	137	4	this	this	DT
3262	137	5	area	area	NN
3262	137	6	include	include	VBP
3262	137	7	Carnegie	Carnegie	NNP
3262	137	8	Mellon	Mellon	NNP
3262	137	9	University	University	NNP
3262	137	10	’s	’s	POS
3262	137	11	Informedia	Informedia	NNP
3262	137	12	Digital	Digital	NNP
3262	137	13	Video	Video	NNP
3262	137	14	Library	Library	NNP
3262	137	15	(	(	-LRB-
3262	137	16	www	www	NNP
3262	137	17	.	.	.
3262	138	1	informedia.cs.cmu.edu	informedia.cs.cmu.edu	NFP
3262	138	2	)	)	-RRB-
3262	138	3	and	and	CC
3262	138	4	MUVIS	MUVIS	NNP
3262	138	5	of	of	IN
3262	138	6	Finland	Finland	NNP
3262	138	7	’s	’s	,
3262	138	8	Tampere	Tampere	NNP
3262	138	9	University	University	NNP
3262	138	10	of	of	IN
3262	138	11	Technology	Technology	NNP
3262	138	12	(	(	-LRB-
3262	138	13	http://muvis.cs.tut.fi/index	http://muvis.cs.tut.fi/index	ADD
3262	138	14	.	.	.
3262	139	1	html	html	NNP
3262	139	2	)	)	-RRB-
3262	139	3	.	.	.
3262	140	1	Content	content	NN
3262	140	2	-	-	HYPH
3262	140	3	based	base	VBN
3262	140	4	information	information	NN
3262	140	5	retrieval	retrieval	NN
3262	140	6	for	for	IN
3262	140	7	other	other	JJ
3262	140	8	digital	digital	JJ
3262	140	9	formats	format	NNS
3262	140	10	With	with	IN
3262	140	11	the	the	DT
3262	140	12	advance	advance	NN
3262	140	13	of	of	IN
3262	140	14	digitization	digitization	NN
3262	140	15	technology	technology	NN
3262	140	16	,	,	,
3262	140	17	the	the	DT
3262	140	18	content	content	NN
3262	140	19	and	and	CC
3262	140	20	formats	format	NNS
3262	140	21	of	of	IN
3262	140	22	digital	digital	JJ
3262	140	23	libraries	library	NNS
3262	140	24	are	be	VBP
3262	140	25	much	much	RB
3262	140	26	richer	rich	JJR
3262	140	27	than	than	IN
3262	140	28	before	before	RB
3262	140	29	.	.	.
3262	141	1	They	-PRON-	PRP
3262	141	2	are	be	VBP
3262	141	3	not	not	RB
3262	141	4	limited	limit	VBN
3262	141	5	to	to	IN
3262	141	6	text	text	NN
3262	141	7	,	,	,
3262	141	8	image	image	NN
3262	141	9	,	,	,
3262	141	10	audio	audio	NN
3262	141	11	,	,	,
3262	141	12	and	and	CC
3262	141	13	video	video	NN
3262	141	14	.	.	.
3262	142	1	Some	some	DT
3262	142	2	new	new	JJ
3262	142	3	formats	format	NNS
3262	142	4	of	of	IN
3262	142	5	digital	digital	JJ
3262	142	6	content	content	NN
3262	142	7	are	be	VBP
3262	142	8	emerging	emerge	VBG
3262	142	9	.	.	.
3262	143	1	Digital	digital	JJ
3262	143	2	libraries	library	NNS
3262	143	3	of	of	IN
3262	143	4	3-D	3-d	CD
3262	143	5	objects	object	NNS
3262	143	6	are	be	VBP
3262	143	7	good	good	JJ
3262	143	8	examples	example	NNS
3262	143	9	.	.	.
3262	144	1	Since	since	IN
3262	144	2	3-D	3-d	CD
3262	144	3	models	model	NNS
3262	144	4	have	have	VBP
3262	144	5	arbitrary	arbitrary	JJ
3262	144	6	topologies	topology	NNS
3262	144	7	and	and	CC
3262	144	8	can-	can-	JJS
3262	144	9	not	not	RB
3262	144	10	be	be	VB
3262	144	11	easily	easily	RB
3262	144	12	“	"	``
3262	144	13	parameterized	parameterize	VBN
3262	144	14	”	"	''
3262	144	15	using	use	VBG
3262	144	16	a	a	DT
3262	144	17	standard	standard	JJ
3262	144	18	template	template	NN
3262	144	19	as	as	IN
3262	144	20	in	in	IN
3262	144	21	the	the	DT
3262	144	22	case	case	NN
3262	144	23	for	for	IN
3262	144	24	2-D	2-d	CD
3262	144	25	forms	form	NNS
3262	144	26	(	(	-LRB-
3262	144	27	Bustos	Bustos	NNP
3262	144	28	et	et	NNP
3262	144	29	al	al	NNP
3262	144	30	.	.	.
3262	145	1	2005	2005	CD
3262	145	2	)	)	-RRB-
3262	145	3	,	,	,
3262	145	4	content-	content-	NNP
3262	145	5	based	base	VBN
3262	145	6	3-D	3-d	CD
3262	145	7	model	model	NN
3262	145	8	retrieval	retrieval	NN
3262	145	9	is	be	VBZ
3262	145	10	a	a	DT
3262	145	11	more	more	RBR
3262	145	12	challenging	challenging	JJ
3262	145	13	research	research	NN
3262	145	14	topic	topic	NN
3262	145	15	than	than	IN
3262	145	16	other	other	JJ
3262	145	17	multimedia	multimedia	NNS
3262	145	18	formats	format	NNS
3262	145	19	discussed	discuss	VBN
3262	145	20	earlier	early	RBR
3262	145	21	.	.	.
3262	146	1	So	so	RB
3262	146	2	far	far	RB
3262	146	3	,	,	,
3262	146	4	four	four	CD
3262	146	5	types	type	NNS
3262	146	6	of	of	IN
3262	146	7	solutions	solution	NNS
3262	146	8	—	—	:
3262	146	9	primitive	primitive	NN
3262	146	10	-	-	HYPH
3262	146	11	based	base	VBN
3262	146	12	,	,	,
3262	146	13	statis-	statis-	NNP
3262	146	14	tics	tics	NNP
3262	146	15	-	-	HYPH
3262	146	16	based	base	VBN
3262	146	17	,	,	,
3262	146	18	geometry	geometry	NN
3262	146	19	-	-	HYPH
3262	146	20	based	base	VBN
3262	146	21	,	,	,
3262	146	22	and	and	CC
3262	146	23	view	view	NN
3262	146	24	-	-	HYPH
3262	146	25	based	base	VBN
3262	146	26	—	—	:
3262	146	27	have	have	VBP
3262	146	28	been	be	VBN
3262	146	29	found	find	VBN
3262	146	30	(	(	-LRB-
3262	146	31	Bimbo	Bimbo	NNP
3262	146	32	and	and	CC
3262	146	33	Pala	Pala	NNP
3262	146	34	2006	2006	CD
3262	146	35	)	)	-RRB-
3262	146	36	.	.	.
3262	147	1	Primitive	primitive	NN
3262	147	2	-	-	HYPH
3262	147	3	based	base	VBN
3262	147	4	solutions	solution	NNS
3262	147	5	represent	represent	VBP
3262	147	6	3-D	3-d	CD
3262	147	7	objects	object	NNS
3262	147	8	with	with	IN
3262	147	9	a	a	DT
3262	147	10	basic	basic	JJ
3262	147	11	set	set	NN
3262	147	12	of	of	IN
3262	147	13	parameterized	parameterize	VBN
3262	147	14	primitive	primitive	JJ
3262	147	15	elements	element	NNS
3262	147	16	.	.	.
3262	148	1	Parameters	parameter	NNS
3262	148	2	are	be	VBP
3262	148	3	used	use	VBN
3262	148	4	to	to	TO
3262	148	5	control	control	VB
3262	148	6	the	the	DT
3262	148	7	shape	shape	NN
3262	148	8	of	of	IN
3262	148	9	each	each	DT
3262	148	10	primitive	primitive	JJ
3262	148	11	element	element	NN
3262	148	12	and	and	CC
3262	148	13	to	to	TO
3262	148	14	fit	fit	VB
3262	148	15	each	each	DT
3262	148	16	primitive	primitive	JJ
3262	148	17	element	element	NN
3262	148	18	with	with	IN
3262	148	19	a	a	DT
3262	148	20	part	part	NN
3262	148	21	of	of	IN
3262	148	22	the	the	DT
3262	148	23	model	model	NN
3262	148	24	.	.	.
3262	149	1	With	with	IN
3262	149	2	statistics	statistic	NNS
3262	149	3	-	-	HYPH
3262	149	4	based	base	VBN
3262	149	5	approaches	approach	NNS
3262	149	6	,	,	,
3262	149	7	shape	shape	NN
3262	149	8	descriptions	description	NNS
3262	149	9	based	base	VBN
3262	149	10	on	on	IN
3262	149	11	statistical	statistical	JJ
3262	149	12	mod-	mod-	NN
3262	149	13	Figure	Figure	NNP
3262	149	14	5.1	5.1	CD
3262	149	15	.	.	.
3262	150	1	The	the	DT
3262	150	2	user	user	NN
3262	150	3	composes	compose	VBZ
3262	150	4	a	a	DT
3262	150	5	query	query	NN
3262	150	6	Figure	figure	NN
3262	150	7	5.2	5.2	CD
3262	150	8	.	.	.
3262	151	1	Search	search	NN
3262	151	2	results	result	NNS
3262	151	3	for	for	IN
3262	151	4	the	the	DT
3262	151	5	sample	sample	NN
3262	151	6	query	query	NN
3262	151	7	46	46	CD
3262	151	8	inFoRmation	information	NN
3262	151	9	tECHnoLoGY	tECHnoLoGY	:
3262	151	10	anD	and	CC
3262	151	11	LiBRaRiEs	LiBRaRiEs	NNP
3262	151	12	|	|	NNP
3262	151	13	maRCH	maRCH	VBZ
3262	151	14	200846	200846	CD
3262	151	15	inFoRmation	information	NN
3262	151	16	tECHnoLoGY	tECHnoLoGY	:
3262	151	17	anD	and	CC
3262	151	18	LiBRaRiEs	LiBRaRiEs	NNP
3262	151	19	|	|	NNP
3262	151	20	maRCH	march	CD
3262	151	21	2008	2008	CD
3262	151	22	els	el	NNS
3262	151	23	are	be	VBP
3262	151	24	created	create	VBN
3262	151	25	and	and	CC
3262	151	26	measured	measure	VBN
3262	151	27	.	.	.
3262	152	1	Geometry	geometry	NN
3262	152	2	-	-	HYPH
3262	152	3	based	base	VBN
3262	152	4	methods	method	NNS
3262	152	5	,	,	,
3262	152	6	however	however	RB
3262	152	7	,	,	,
3262	152	8	use	use	VB
3262	152	9	geometric	geometric	JJ
3262	152	10	properties	property	NNS
3262	152	11	of	of	IN
3262	152	12	the	the	DT
3262	152	13	3-D	3-d	CD
3262	152	14	object	object	NN
3262	152	15	and	and	CC
3262	152	16	their	-PRON-	PRP$
3262	152	17	measures	measure	NNS
3262	152	18	as	as	IN
3262	152	19	global	global	JJ
3262	152	20	shape	shape	NN
3262	152	21	descriptors	descriptor	NNS
3262	152	22	.	.	.
3262	153	1	For	for	IN
3262	153	2	view-	view-	JJ
3262	153	3	based	base	VBN
3262	153	4	solutions	solution	NNS
3262	153	5	,	,	,
3262	153	6	a	a	DT
3262	153	7	set	set	NN
3262	153	8	of	of	IN
3262	153	9	2-D	2-d	CD
3262	153	10	views	view	NNS
3262	153	11	of	of	IN
3262	153	12	the	the	DT
3262	153	13	model	model	NN
3262	153	14	and	and	CC
3262	153	15	descriptors	descriptor	NNS
3262	153	16	of	of	IN
3262	153	17	their	-PRON-	PRP$
3262	153	18	content	content	NN
3262	153	19	are	be	VBP
3262	153	20	used	use	VBN
3262	153	21	to	to	TO
3262	153	22	represent	represent	VB
3262	153	23	the	the	DT
3262	153	24	3-D	3-d	CD
3262	153	25	object	object	NN
3262	153	26	shape	shape	NN
3262	153	27	(	(	-LRB-
3262	153	28	Bimbo	Bimbo	NNP
3262	153	29	and	and	CC
3262	153	30	Pala	Pala	NNP
3262	153	31	2006	2006	CD
3262	153	32	)	)	-RRB-
3262	153	33	.	.	.
3262	154	1	Another	another	DT
3262	154	2	novel	novel	JJ
3262	154	3	example	example	NN
3262	154	4	is	be	VBZ
3262	154	5	Moustakas	Moustakas	NNP
3262	154	6	et	et	FW
3262	154	7	al	al	NNP
3262	154	8	.	.	.
3262	154	9	’s	’s	''
3262	154	10	(	(	-LRB-
3262	154	11	2005	2005	CD
3262	154	12	)	)	-RRB-
3262	154	13	project	project	NN
3262	154	14	on	on	IN
3262	154	15	3-D	3-d	CD
3262	154	16	model	model	NN
3262	154	17	search	search	NN
3262	154	18	using	use	VBG
3262	154	19	sketches	sketch	NNS
3262	154	20	.	.	.
3262	155	1	In	in	IN
3262	155	2	the	the	DT
3262	155	3	experimental	experimental	JJ
3262	155	4	system	system	NN
3262	155	5	,	,	,
3262	155	6	the	the	DT
3262	155	7	vector	vector	NN
3262	155	8	of	of	IN
3262	155	9	geometrical	geometrical	JJ
3262	155	10	descrip-	descrip-	XX
3262	155	11	tors	tor	NNS
3262	155	12	for	for	IN
3262	155	13	each	each	DT
3262	155	14	3-D	3-d	CD
3262	155	15	model	model	NN
3262	155	16	is	be	VBZ
3262	155	17	calculated	calculate	VBN
3262	155	18	during	during	IN
3262	155	19	the	the	DT
3262	155	20	feature	feature	NN
3262	155	21	extraction	extraction	NN
3262	155	22	stage	stage	NN
3262	155	23	.	.	.
3262	156	1	In	in	IN
3262	156	2	the	the	DT
3262	156	3	retrieval	retrieval	NN
3262	156	4	stage	stage	NN
3262	156	5	,	,	,
3262	156	6	a	a	DT
3262	156	7	user	user	NN
3262	156	8	can	can	MD
3262	156	9	ini-	ini-	VB
3262	156	10	tially	tially	RB
3262	156	11	use	use	VB
3262	156	12	one	one	CD
3262	156	13	of	of	IN
3262	156	14	the	the	DT
3262	156	15	sketching	sketch	VBG
3262	156	16	interfaces	interface	NNS
3262	156	17	(	(	-LRB-
3262	156	18	such	such	JJ
3262	156	19	as	as	IN
3262	156	20	the	the	DT
3262	156	21	virtual	virtual	JJ
3262	156	22	reality	reality	NN
3262	156	23	interface	interface	NN
3262	156	24	or	or	CC
3262	156	25	by	by	IN
3262	156	26	using	use	VBG
3262	156	27	an	an	DT
3262	156	28	air	air	NN
3262	156	29	mouse	mouse	NN
3262	156	30	)	)	-RRB-
3262	156	31	to	to	TO
3262	156	32	sketch	sketch	VB
3262	156	33	a	a	DT
3262	156	34	2-D	2-d	CD
3262	156	35	contour	contour	NN
3262	156	36	of	of	IN
3262	156	37	the	the	DT
3262	156	38	desired	desire	VBN
3262	156	39	3-D	3-d	CD
3262	156	40	object	object	NN
3262	156	41	.	.	.
3262	157	1	The	the	DT
3262	157	2	2-D	2-d	CD
3262	157	3	shape	shape	NN
3262	157	4	is	be	VBZ
3262	157	5	recognized	recognize	VBN
3262	157	6	by	by	IN
3262	157	7	the	the	DT
3262	157	8	system	system	NN
3262	157	9	,	,	,
3262	157	10	and	and	CC
3262	157	11	a	a	DT
3262	157	12	sample	sample	NN
3262	157	13	primi-	primi-	NN
3262	157	14	tive	tive	JJ
3262	157	15	is	be	VBZ
3262	157	16	automatically	automatically	RB
3262	157	17	inserted	insert	VBN
3262	157	18	in	in	IN
3262	157	19	the	the	DT
3262	157	20	scene	scene	NN
3262	157	21	.	.	.
3262	158	1	Next	next	RB
3262	158	2	,	,	,
3262	158	3	the	the	DT
3262	158	4	user	user	NN
3262	158	5	defines	define	VBZ
3262	158	6	other	other	JJ
3262	158	7	elements	element	NNS
3262	158	8	that	that	WDT
3262	158	9	can	can	MD
3262	158	10	not	not	RB
3262	158	11	be	be	VB
3262	158	12	described	describe	VBN
3262	158	13	by	by	IN
3262	158	14	the	the	DT
3262	158	15	2-D	2-d	CD
3262	158	16	contour	contour	NN
3262	158	17	,	,	,
3262	158	18	such	such	JJ
3262	158	19	as	as	IN
3262	158	20	the	the	DT
3262	158	21	height	height	NN
3262	158	22	of	of	IN
3262	158	23	the	the	DT
3262	158	24	object	object	NN
3262	158	25	,	,	,
3262	158	26	and	and	CC
3262	158	27	manip-	manip-	JJ
3262	158	28	ulates	ulate	NNS
3262	158	29	the	the	DT
3262	158	30	2-D	2-d	CD
3262	158	31	contour	contour	NN
3262	158	32	until	until	IN
3262	158	33	it	-PRON-	PRP
3262	158	34	reaches	reach	VBZ
3262	158	35	its	-PRON-	PRP$
3262	158	36	target	target	NN
3262	158	37	position	position	NN
3262	158	38	.	.	.
3262	159	1	The	the	DT
3262	159	2	final	final	JJ
3262	159	3	query	query	NN
3262	159	4	is	be	VBZ
3262	159	5	formed	form	VBN
3262	159	6	after	after	IN
3262	159	7	all	all	PDT
3262	159	8	the	the	DT
3262	159	9	primitives	primitive	NNS
3262	159	10	are	be	VBP
3262	159	11	inserted	insert	VBN
3262	159	12	.	.	.
3262	160	1	Finally	finally	RB
3262	160	2	,	,	,
3262	160	3	the	the	DT
3262	160	4	system	system	NN
3262	160	5	computes	compute	VBZ
3262	160	6	the	the	DT
3262	160	7	similarities	similarity	NNS
3262	160	8	between	between	IN
3262	160	9	the	the	DT
3262	160	10	query	query	NN
3262	160	11	model	model	NN
3262	160	12	and	and	CC
3262	160	13	each	each	DT
3262	160	14	3-D	3-d	CD
3262	160	15	model	model	NN
3262	160	16	in	in	IN
3262	160	17	the	the	DT
3262	160	18	database	database	NN
3262	160	19	,	,	,
3262	160	20	and	and	CC
3262	160	21	renders	render	VBZ
3262	160	22	the	the	DT
3262	160	23	best	good	JJS
3262	160	24	matching	matching	NN
3262	160	25	records	record	NNS
3262	160	26	.	.	.
3262	161	1	An	an	DT
3262	161	2	online	online	JJ
3262	161	3	demonstration	demonstration	NN
3262	161	4	can	can	MD
3262	161	5	be	be	VB
3262	161	6	found	find	VBN
3262	161	7	for	for	IN
3262	161	8	a	a	DT
3262	161	9	European	european	JJ
3262	161	10	project	project	NN
3262	161	11	specifically	specifically	RB
3262	161	12	designed	design	VBN
3262	161	13	for	for	IN
3262	161	14	a	a	DT
3262	161	15	3-D	3-d	CD
3262	161	16	digital	digital	NNP
3262	161	17	museum	museum	NN
3262	161	18	col-	col-	NN
3262	161	19	lection	lection	NN
3262	161	20	,	,	,
3262	161	21	SCULPTEUR	SCULPTEUR	NNP
3262	161	22	(	(	-LRB-
3262	161	23	www.sculpteurweb.org	www.sculpteurweb.org	NN
3262	161	24	)	)	-RRB-
3262	161	25	.	.	.
3262	162	1	From	from	IN
3262	162	2	its	-PRON-	PRP$
3262	162	3	Web	web	NN
3262	162	4	-	-	HYPH
3262	162	5	based	base	VBN
3262	162	6	search	search	NN
3262	162	7	interface	interface	NN
3262	162	8	,	,	,
3262	162	9	a	a	DT
3262	162	10	user	user	NN
3262	162	11	can	can	MD
3262	162	12	choose	choose	VB
3262	162	13	to	to	TO
3262	162	14	do	do	VB
3262	162	15	a	a	DT
3262	162	16	meta-	meta-	JJ
3262	162	17	data	data	NN
3262	162	18	search	search	NN
3262	162	19	or	or	CC
3262	162	20	content	content	NN
3262	162	21	-	-	HYPH
3262	162	22	based	base	VBN
3262	162	23	search	search	NN
3262	162	24	for	for	IN
3262	162	25	a	a	DT
3262	162	26	3-D	3-d	CD
3262	162	27	object	object	NN
3262	162	28	.	.	.
3262	163	1	The	the	DT
3262	163	2	search	search	NN
3262	163	3	strategy	strategy	NN
3262	163	4	here	here	RB
3262	163	5	is	be	VBZ
3262	163	6	somewhat	somewhat	RB
3262	163	7	similar	similar	JJ
3262	163	8	to	to	IN
3262	163	9	that	that	DT
3262	163	10	in	in	IN
3262	163	11	some	some	DT
3262	163	12	CBIR	CBIR	NNP
3262	163	13	systems	system	NNS
3262	163	14	:	:	:
3262	163	15	the	the	DT
3262	163	16	user	user	NN
3262	163	17	can	can	MD
3262	163	18	upload	upload	VB
3262	163	19	a	a	DT
3262	163	20	3-D	3-d	CD
3262	163	21	model	model	NN
3262	163	22	in	in	IN
3262	163	23	VRML	VRML	NNP
3262	163	24	formats	format	NNS
3262	163	25	,	,	,
3262	163	26	then	then	RB
3262	163	27	select	select	VB
3262	163	28	a	a	DT
3262	163	29	search	search	NN
3262	163	30	algorithm	algorithm	NN
3262	163	31	(	(	-LRB-
3262	163	32	such	such	JJ
3262	163	33	as	as	IN
3262	163	34	similar	similar	JJ
3262	163	35	color	color	NN
3262	163	36	,	,	,
3262	163	37	texture	texture	NN
3262	163	38	,	,	,
3262	163	39	etc	etc	FW
3262	163	40	.	.	.
3262	163	41	)	)	-RRB-
3262	164	1	to	to	TO
3262	164	2	perform	perform	VB
3262	164	3	a	a	DT
3262	164	4	search	search	NN
3262	164	5	within	within	IN
3262	164	6	a	a	DT
3262	164	7	digital	digital	JJ
3262	164	8	collection	collection	NN
3262	164	9	of	of	IN
3262	164	10	3-D	3-d	CD
3262	164	11	models	model	NNS
3262	164	12	.	.	.
3262	165	1	As	as	IN
3262	165	2	3-D	3-d	CD
3262	165	3	computer	computer	NN
3262	165	4	visualization	visualization	NN
3262	165	5	has	have	VBZ
3262	165	6	been	be	VBN
3262	165	7	widely	widely	RB
3262	165	8	used	use	VBN
3262	165	9	in	in	IN
3262	165	10	a	a	DT
3262	165	11	variety	variety	NN
3262	165	12	of	of	IN
3262	165	13	areas	area	NNS
3262	165	14	,	,	,
3262	165	15	there	there	EX
3262	165	16	are	be	VBP
3262	165	17	more	more	JJR
3262	165	18	research	research	NN
3262	165	19	projects	project	NNS
3262	165	20	focusing	focus	VBG
3262	165	21	on	on	IN
3262	165	22	the	the	DT
3262	165	23	content	content	NN
3262	165	24	-	-	HYPH
3262	165	25	based	base	VBN
3262	165	26	informa-	informa-	JJ
3262	165	27	tion	tion	NN
3262	165	28	retrieval	retrieval	NN
3262	165	29	techniques	technique	NNS
3262	165	30	for	for	IN
3262	165	31	this	this	DT
3262	165	32	new	new	JJ
3262	165	33	multimedia	multimedia	NNS
3262	165	34	format	format	NN
3262	165	35	.	.	.
3262	166	1	■	■	NFP
3262	166	2	Conclusion	Conclusion	NNP
3262	166	3	There	there	EX
3262	166	4	is	be	VBZ
3262	166	5	no	no	DT
3262	166	6	doubt	doubt	NN
3262	166	7	that	that	IN
3262	166	8	content	content	NN
3262	166	9	-	-	HYPH
3262	166	10	based	base	VBN
3262	166	11	information	information	NN
3262	166	12	retrieval	retrieval	NN
3262	166	13	technology	technology	NN
3262	166	14	is	be	VBZ
3262	166	15	an	an	DT
3262	166	16	emerging	emerge	VBG
3262	166	17	trend	trend	NN
3262	166	18	for	for	IN
3262	166	19	digital	digital	JJ
3262	166	20	library	library	NN
3262	166	21	development	development	NN
3262	166	22	and	and	CC
3262	166	23	will	will	MD
3262	166	24	be	be	VB
3262	166	25	an	an	DT
3262	166	26	important	important	JJ
3262	166	27	comple-	comple-	NN
3262	166	28	ment	ment	JJ
3262	166	29	to	to	IN
3262	166	30	the	the	DT
3262	166	31	traditional	traditional	JJ
3262	166	32	text	text	NN
3262	166	33	-	-	HYPH
3262	166	34	based	base	VBN
3262	166	35	retrieval	retrieval	NN
3262	166	36	technology	technology	NN
3262	166	37	.	.	.
3262	167	1	The	the	DT
3262	167	2	ideal	ideal	JJ
3262	167	3	CBIR	CBIR	NNP
3262	167	4	system	system	NN
3262	167	5	can	can	MD
3262	167	6	semantically	semantically	RB
3262	167	7	understand	understand	VB
3262	167	8	the	the	DT
3262	167	9	information	information	NN
3262	167	10	in	in	IN
3262	167	11	a	a	DT
3262	167	12	digital	digital	JJ
3262	167	13	library	library	NN
3262	167	14	,	,	,
3262	167	15	and	and	CC
3262	167	16	render	render	VB
3262	167	17	users	user	NNS
3262	167	18	the	the	DT
3262	167	19	most	most	RBS
3262	167	20	desirable	desirable	JJ
3262	167	21	data	datum	NNS
3262	167	22	.	.	.
3262	168	1	However	however	RB
3262	168	2	,	,	,
3262	168	3	the	the	DT
3262	168	4	machine	machine	NN
3262	168	5	understand-	understand-	NNP
3262	168	6	ing	ing	NNP
3262	168	7	of	of	IN
3262	168	8	semantic	semantic	JJ
3262	168	9	information	information	NN
3262	168	10	still	still	RB
3262	168	11	remains	remain	VBZ
3262	168	12	to	to	TO
3262	168	13	be	be	VB
3262	168	14	a	a	DT
3262	168	15	great	great	JJ
3262	168	16	difficulty	difficulty	NN
3262	168	17	.	.	.
3262	169	1	Therefore	therefore	RB
3262	169	2	,	,	,
3262	169	3	most	most	JJS
3262	169	4	current	current	JJ
3262	169	5	research	research	NN
3262	169	6	projects	project	NNS
3262	169	7	,	,	,
3262	169	8	including	include	VBG
3262	169	9	those	those	DT
3262	169	10	discussed	discuss	VBN
3262	169	11	in	in	IN
3262	169	12	this	this	DT
3262	169	13	paper	paper	NN
3262	169	14	,	,	,
3262	169	15	deal	deal	VBP
3262	169	16	with	with	IN
3262	169	17	the	the	DT
3262	169	18	understanding	understanding	NN
3262	169	19	and	and	CC
3262	169	20	retrieval	retrieval	NN
3262	169	21	of	of	IN
3262	169	22	lower	low	JJR
3262	169	23	-	-	HYPH
3262	169	24	level	level	NN
3262	169	25	features	feature	NNS
3262	169	26	or	or	CC
3262	169	27	physical	physical	JJ
3262	169	28	features	feature	NNS
3262	169	29	of	of	IN
3262	169	30	multimedia	multimedia	NNS
3262	169	31	content	content	NN
3262	169	32	.	.	.
3262	170	1	Certainly	certainly	RB
3262	170	2	,	,	,
3262	170	3	as	as	IN
3262	170	4	related	related	JJ
3262	170	5	disciplines	discipline	NNS
3262	170	6	such	such	JJ
3262	170	7	as	as	IN
3262	170	8	computer	computer	NN
3262	170	9	vision	vision	NN
3262	170	10	and	and	CC
3262	170	11	artificial	artificial	JJ
3262	170	12	intelligence	intelligence	NN
3262	170	13	keep	keep	VBP
3262	170	14	developing	develop	VBG
3262	170	15	,	,	,
3262	170	16	more	more	JJR
3262	170	17	researches	research	NNS
3262	170	18	will	will	MD
3262	170	19	be	be	VB
3262	170	20	done	do	VBN
3262	170	21	on	on	IN
3262	170	22	higher	high	JJR
3262	170	23	-	-	HYPH
3262	170	24	level	level	NN
3262	170	25	feature	feature	NN
3262	170	26	-	-	HYPH
3262	170	27	based	base	VBN
3262	170	28	retrieval	retrieval	NN
3262	170	29	.	.	.
3262	171	1	In	in	IN
3262	171	2	addition	addition	NN
3262	171	3	,	,	,
3262	171	4	the	the	DT
3262	171	5	growing	grow	VBG
3262	171	6	varieties	variety	NNS
3262	171	7	of	of	IN
3262	171	8	multimedia	multimedia	NNS
3262	171	9	content	content	NN
3262	171	10	in	in	IN
3262	171	11	digital	digital	JJ
3262	171	12	libraries	library	NNS
3262	171	13	have	have	VBP
3262	171	14	also	also	RB
3262	171	15	brought	bring	VBN
3262	171	16	many	many	JJ
3262	171	17	new	new	JJ
3262	171	18	challenges	challenge	NNS
3262	171	19	.	.	.
3262	172	1	For	for	IN
3262	172	2	instance	instance	NN
3262	172	3	,	,	,
3262	172	4	3-D	3-d	CD
3262	172	5	models	model	NNS
3262	172	6	now	now	RB
3262	172	7	become	become	VBP
3262	172	8	impor-	impor-	JJ
3262	172	9	tant	tant	JJ
3262	172	10	components	component	NNS
3262	172	11	of	of	IN
3262	172	12	many	many	JJ
3262	172	13	digital	digital	JJ
3262	172	14	libraries	library	NNS
3262	172	15	and	and	CC
3262	172	16	museums	museum	NNS
3262	172	17	.	.	.
3262	173	1	Content	content	NN
3262	173	2	-	-	HYPH
3262	173	3	based	base	VBN
3262	173	4	retrieval	retrieval	NN
3262	173	5	technology	technology	NN
3262	173	6	can	can	MD
3262	173	7	be	be	VB
3262	173	8	a	a	DT
3262	173	9	good	good	JJ
3262	173	10	direc-	direc-	NN
3262	173	11	tion	tion	NN
3262	173	12	for	for	IN
3262	173	13	this	this	DT
3262	173	14	type	type	NN
3262	173	15	of	of	IN
3262	173	16	content	content	NN
3262	173	17	,	,	,
3262	173	18	since	since	IN
3262	173	19	the	the	DT
3262	173	20	shapes	shape	NNS
3262	173	21	of	of	IN
3262	173	22	these	these	DT
3262	173	23	3-D	3-d	CD
3262	173	24	objects	object	NNS
3262	173	25	are	be	VBP
3262	173	26	often	often	RB
3262	173	27	found	find	VBN
3262	173	28	more	more	RBR
3262	173	29	effectively	effectively	RB
3262	173	30	if	if	IN
3262	173	31	the	the	DT
3262	173	32	user	user	NN
3262	173	33	can	can	MD
3262	173	34	compose	compose	VB
3262	173	35	the	the	DT
3262	173	36	query	query	NN
3262	173	37	visually	visually	RB
3262	173	38	.	.	.
3262	174	1	New	new	JJ
3262	174	2	CBIR	CBIR	NNP
3262	174	3	approaches	approach	NNS
3262	174	4	need	need	VBP
3262	174	5	to	to	TO
3262	174	6	be	be	VB
3262	174	7	developed	develop	VBN
3262	174	8	for	for	IN
3262	174	9	these	these	DT
3262	174	10	novel	novel	JJ
3262	174	11	formats	format	NNS
3262	174	12	.	.	.
3262	175	1	Furthermore	furthermore	RB
3262	175	2	,	,	,
3262	175	3	most	most	JJS
3262	175	4	CBIR	CBIR	NNP
3262	175	5	projects	project	NNS
3262	175	6	today	today	NN
3262	175	7	tend	tend	VBP
3262	175	8	to	to	TO
3262	175	9	be	be	VB
3262	175	10	Web	web	NN
3262	175	11	-	-	HYPH
3262	175	12	based	base	VBN
3262	175	13	.	.	.
3262	176	1	By	by	IN
3262	176	2	contrast	contrast	NN
3262	176	3	,	,	,
3262	176	4	many	many	JJ
3262	176	5	project	project	NN
3262	176	6	were	be	VBD
3262	176	7	based	base	VBN
3262	176	8	on	on	IN
3262	176	9	client	client	NN
3262	176	10	applications	application	NNS
3262	176	11	in	in	IN
3262	176	12	the	the	DT
3262	176	13	1990s	1990s	CD
3262	176	14	.	.	.
3262	177	1	These	these	DT
3262	177	2	Web	web	NN
3262	177	3	-	-	HYPH
3262	177	4	based	base	VBN
3262	177	5	CBIR	CBIR	NNP
3262	177	6	tools	tool	NNS
3262	177	7	will	will	MD
3262	177	8	have	have	VB
3262	177	9	significant	significant	JJ
3262	177	10	influence	influence	NN
3262	177	11	on	on	IN
3262	177	12	digital	digital	JJ
3262	177	13	libraries	library	NNS
3262	177	14	or	or	CC
3262	177	15	repositories	repository	NNS
3262	177	16	,	,	,
3262	177	17	as	as	IN
3262	177	18	most	most	JJS
3262	177	19	of	of	IN
3262	177	20	them	-PRON-	PRP
3262	177	21	are	be	VBP
3262	177	22	also	also	RB
3262	177	23	Web	web	NN
3262	177	24	-	-	HYPH
3262	177	25	based	base	VBN
3262	177	26	.	.	.
3262	178	1	Particularly	particularly	RB
3262	178	2	in	in	IN
3262	178	3	the	the	DT
3262	178	4	age	age	NN
3262	178	5	of	of	IN
3262	178	6	Web	web	NN
3262	178	7	2.0	2.0	CD
3262	178	8	,	,	,
3262	178	9	some	some	DT
3262	178	10	large	large	JJ
3262	178	11	digital	digital	JJ
3262	178	12	repositories	repository	NNS
3262	178	13	—	—	:
3262	178	14	such	such	JJ
3262	178	15	as	as	IN
3262	178	16	Flickr	Flickr	NNP
3262	178	17	for	for	IN
3262	178	18	images	image	NNS
3262	178	19	and	and	CC
3262	178	20	YouTube	YouTube	NNP
3262	178	21	and	and	CC
3262	178	22	Google	Google	NNP
3262	178	23	Video	Video	NNP
3262	178	24	for	for	IN
3262	178	25	video	video	NN
3262	178	26	—	—	:
3262	178	27	are	be	VBP
3262	178	28	changing	change	VBG
3262	178	29	people	people	NNS
3262	178	30	’s	’s	POS
3262	178	31	daily	daily	JJ
3262	178	32	lives	life	NNS
3262	178	33	.	.	.
3262	179	1	The	the	DT
3262	179	2	implementation	implementation	NN
3262	179	3	of	of	IN
3262	179	4	CBIR	CBIR	NNP
3262	179	5	will	will	MD
3262	179	6	be	be	VB
3262	179	7	a	a	DT
3262	179	8	great	great	JJ
3262	179	9	benefit	benefit	NN
3262	179	10	to	to	IN
3262	179	11	millions	million	NNS
3262	179	12	of	of	IN
3262	179	13	users	user	NNS
3262	179	14	.	.	.
3262	180	1	Since	since	IN
3262	180	2	the	the	DT
3262	180	3	nature	nature	NN
3262	180	4	of	of	IN
3262	180	5	CBIR	CBIR	NNP
3262	180	6	is	be	VBZ
3262	180	7	to	to	TO
3262	180	8	provide	provide	VB
3262	180	9	better	well	JJR
3262	180	10	search	search	NN
3262	180	11	aids	aid	NNS
3262	180	12	to	to	TO
3262	180	13	end	end	VB
3262	180	14	users	user	NNS
3262	180	15	,	,	,
3262	180	16	it	-PRON-	PRP
3262	180	17	is	be	VBZ
3262	180	18	extremely	extremely	RB
3262	180	19	important	important	JJ
3262	180	20	to	to	TO
3262	180	21	focus	focus	VB
3262	180	22	on	on	IN
3262	180	23	the	the	DT
3262	180	24	actual	actual	JJ
3262	180	25	user	user	NN
3262	180	26	’s	’s	POS
3262	180	27	needs	need	VBZ
3262	180	28	and	and	CC
3262	180	29	how	how	WRB
3262	180	30	well	well	RB
3262	180	31	the	the	DT
3262	180	32	user	user	NN
3262	180	33	can	can	MD
3262	180	34	use	use	VB
3262	180	35	these	these	DT
3262	180	36	new	new	JJ
3262	180	37	search	search	NN
3262	180	38	tools	tool	NNS
3262	180	39	.	.	.
3262	181	1	It	-PRON-	PRP
3262	181	2	is	be	VBZ
3262	181	3	surprising	surprising	JJ
3262	181	4	to	to	TO
3262	181	5	find	find	VB
3262	181	6	that	that	DT
3262	181	7	little	little	JJ
3262	181	8	usabil-	usabil-	JJ
3262	181	9	ity	ity	NN
3262	181	10	testing	testing	NN
3262	181	11	has	have	VBZ
3262	181	12	been	be	VBN
3262	181	13	done	do	VBN
3262	181	14	for	for	IN
3262	181	15	most	most	JJS
3262	181	16	CBIR	cbir	JJ
3262	181	17	projects	project	NNS
3262	181	18	.	.	.
3262	182	1	Such	such	JJ
3262	182	2	testing	testing	NN
3262	182	3	should	should	MD
3262	182	4	be	be	VB
3262	182	5	incorporated	incorporate	VBN
3262	182	6	into	into	IN
3262	182	7	future	future	JJ
3262	182	8	CBIR	CBIR	NNP
3262	182	9	research	research	NN
3262	182	10	before	before	IN
3262	182	11	it	-PRON-	PRP
3262	182	12	is	be	VBZ
3262	182	13	widely	widely	RB
3262	182	14	adopted	adopt	VBN
3262	182	15	.	.	.
3262	183	1	Bibliography	Bibliography	NNP
3262	183	2	Bimbo	Bimbo	NNP
3262	183	3	,	,	,
3262	183	4	A.	A.	NNP
3262	183	5	and	and	CC
3262	183	6	P.	P.	NNP
3262	183	7	Pala	Pala	NNP
3262	183	8	.	.	.
3262	184	1	2006	2006	CD
3262	184	2	.	.	.
3262	185	1	Content	content	NN
3262	185	2	-	-	HYPH
3262	185	3	based	base	VBN
3262	185	4	retrieval	retrieval	NN
3262	185	5	of	of	IN
3262	185	6	3-D	3-d	CD
3262	185	7	mod-	mod-	NN
3262	185	8	els	els	NNP
3262	185	9	.	.	.
3262	186	1	ACM	ACM	NNP
3262	186	2	Transactions	Transactions	NNPS
3262	186	3	on	on	IN
3262	186	4	Multimedia	Multimedia	NNP
3262	186	5	Computing	Computing	NNP
3262	186	6	,	,	,
3262	186	7	Communica-	Communica-	NNP
3262	186	8	tions	tion	NNS
3262	186	9	,	,	,
3262	186	10	and	and	CC
3262	186	11	Applications	Applications	NNP
3262	186	12	2	2	CD
3262	186	13	,	,	,
3262	186	14	no	no	UH
3262	186	15	.	.	.
3262	187	1	1	1	LS
3262	187	2	:	:	:
3262	187	3	20–43	20–43	CD
3262	187	4	.	.	.
3262	188	1	Bustos	Bustos	NNP
3262	188	2	,	,	,
3262	188	3	B.	B.	NNP
3262	188	4	,	,	,
3262	188	5	et	et	NNP
3262	188	6	al	al	NNP
3262	188	7	.	.	.
3262	189	1	2005	2005	CD
3262	189	2	.	.	.
3262	190	1	Feature	feature	NN
3262	190	2	-	-	HYPH
3262	190	3	based	base	VBN
3262	190	4	similarity	similarity	NN
3262	190	5	search	search	NN
3262	190	6	in	in	IN
3262	190	7	3-D	3-d	CD
3262	190	8	object	object	NN
3262	190	9	databases	database	NNS
3262	190	10	.	.	.
3262	191	1	ACM	ACM	NNP
3262	191	2	Computing	Computing	NNP
3262	191	3	Surveys	Surveys	NNP
3262	191	4	37	37	CD
3262	191	5	,	,	,
3262	191	6	no	no	UH
3262	191	7	.	.	.
3262	192	1	4	4	LS
3262	192	2	:	:	:
3262	192	3	345–387	345–387	CD
3262	192	4	.	.	.
3262	193	1	Chang	Chang	NNP
3262	193	2	,	,	,
3262	193	3	S.	S.	NNP
3262	193	4	,	,	,
3262	193	5	et	et	NNP
3262	193	6	al	al	NNP
3262	193	7	.	.	.
3262	194	1	1997	1997	CD
3262	194	2	)	)	-RRB-
3262	194	3	.	.	.
3262	195	1	VideoQ	VideoQ	NNP
3262	195	2	:	:	:
3262	195	3	an	an	DT
3262	195	4	automated	automate	VBN
3262	195	5	content	content	NN
3262	195	6	based	base	VBN
3262	195	7	video	video	NN
3262	195	8	search	search	NN
3262	195	9	system	system	NN
3262	195	10	using	use	VBG
3262	195	11	visual	visual	JJ
3262	195	12	cues	cue	NNS
3262	195	13	.	.	.
3262	196	1	In	in	IN
3262	196	2	Proceedings	proceeding	NNS
3262	196	3	of	of	IN
3262	196	4	the	the	DT
3262	196	5	5th	5th	JJ
3262	196	6	ACM	ACM	NNP
3262	196	7	International	International	NNP
3262	196	8	Conference	Conference	NNP
3262	196	9	on	on	IN
3262	196	10	Multimedia	Multimedia	NNP
3262	196	11	,	,	,
3262	196	12	E.	E.	NNP
3262	196	13	P.	P.	NNP
3262	196	14	Glinert	Glinert	NNP
3262	196	15	,	,	,
3262	196	16	et	et	NNP
3262	196	17	al	al	NNP
3262	196	18	.	.	NNP
3262	196	19	,	,	,
3262	196	20	eds	eds	NNP
3262	196	21	.	.	.
3262	197	1	New	New	NNP
3262	197	2	York	York	NNP
3262	197	3	:	:	:
3262	197	4	ACM	ACM	NNP
3262	197	5	.	.	.
3262	198	1	Datta	Datta	NNP
3262	198	2	R.	R.	NNP
3262	198	3	,	,	,
3262	198	4	et	et	NNP
3262	198	5	al	al	NNP
3262	198	6	.	.	.
3262	199	1	2005	2005	CD
3262	199	2	.	.	.
3262	200	1	Content	content	NN
3262	200	2	-	-	HYPH
3262	200	3	based	base	VBN
3262	200	4	image	image	NN
3262	200	5	retrieval	retrieval	NN
3262	200	6	:	:	:
3262	200	7	approaches	approach	NNS
3262	200	8	and	and	CC
3262	200	9	trends	trend	NNS
3262	200	10	of	of	IN
3262	200	11	the	the	DT
3262	200	12	new	new	JJ
3262	200	13	age	age	NN
3262	200	14	.	.	.
3262	201	1	In	in	IN
3262	201	2	Proceedings	proceeding	NNS
3262	201	3	of	of	IN
3262	201	4	the	the	DT
3262	201	5	7th	7th	JJ
3262	201	6	Interna-	Interna-	NNP
3262	201	7	tional	tional	JJ
3262	201	8	Workshop	Workshop	NNP
3262	201	9	on	on	IN
3262	201	10	Multimedia	Multimedia	NNP
3262	201	11	Information	Information	NNP
3262	201	12	Retrieval	Retrieval	NNP
3262	201	13	,	,	,
3262	201	14	in	in	IN
3262	201	15	Con-	Con-	NNP
3262	201	16	junction	junction	NN
3262	201	17	with	with	IN
3262	201	18	ACM	ACM	NNP
3262	201	19	International	International	NNP
3262	201	20	Conference	Conference	NNP
3262	201	21	on	on	IN
3262	201	22	Multimedia	Multimedia	NNP
3262	201	23	,	,	,
3262	201	24	H.	H.	NNP
3262	201	25	Zhang	Zhang	NNP
3262	201	26	,	,	,
3262	201	27	,	,	,
3262	201	28	J.	J.	NNP
3262	201	29	Smith	Smith	NNP
3262	201	30	,	,	,
3262	201	31	and	and	CC
3262	201	32	Q.	Q.	NNP
3262	201	33	Tian	Tian	NNP
3262	201	34	,	,	,
3262	201	35	eds	eds	XX
3262	201	36	.	.	.
3262	202	1	New	New	NNP
3262	202	2	York	York	NNP
3262	202	3	:	:	:
3262	202	4	ACM	ACM	NNP
3262	202	5	.	.	.
3262	203	1	Ellis	Ellis	NNP
3262	203	2	,	,	,
3262	203	3	D.	D.	NNP
3262	203	4	and	and	CC
3262	203	5	K.	K.	NNP
3262	203	6	Lee	Lee	NNP
3262	203	7	.	.	.
3262	204	1	Minimal	minimal	JJ
3262	204	2	-	-	HYPH
3262	204	3	impact	impact	NN
3262	204	4	audio	audio	NN
3262	204	5	-	-	HYPH
3262	204	6	based	base	VBN
3262	204	7	personal	personal	JJ
3262	204	8	archives	archive	NNS
3262	204	9	.	.	.
3262	205	1	In	in	IN
3262	205	2	Proceedings	Proceedings	NNP
3262	205	3	of	of	IN
3262	205	4	the	the	DT
3262	205	5	1st	1st	JJ
3262	205	6	ACM	ACM	NNP
3262	205	7	workshop	workshop	NN
3262	205	8	on	on	IN
3262	205	9	Continuous	Continuous	NNP
3262	205	10	Archival	Archival	NNP
3262	205	11	and	and	CC
3262	205	12	Retrieval	Retrieval	NNP
3262	205	13	of	of	IN
3262	205	14	Personal	Personal	NNP
3262	205	15	Experiences	Experiences	NNPS
3262	205	16	CARPE	CARPE	NNP
3262	205	17	,	,	,
3262	205	18	J.	J.	NNP
3262	205	19	Gem-	Gem-	NNP
3262	205	20	mell	mell	NNP
3262	205	21	,	,	,
3262	205	22	et	et	NNP
3262	205	23	al	al	NNP
3262	205	24	.	.	NNP
3262	205	25	,	,	,
3262	205	26	eds	eds	NNP
3262	205	27	.	.	.
3262	206	1	New	New	NNP
3262	206	2	York	York	NNP
3262	206	3	:	:	:
3262	206	4	ACM	ACM	NNP
3262	206	5	.	.	.
3262	207	1	Li	Li	NNP
3262	207	2	,	,	,
3262	207	3	T.	T.	NNP
3262	207	4	,	,	,
3262	207	5	et	et	NNP
3262	207	6	al	al	NNP
3262	207	7	.	.	.
3262	208	1	2003	2003	CD
3262	208	2	.	.	.
3262	209	1	A	a	DT
3262	209	2	comparative	comparative	JJ
3262	209	3	study	study	NN
3262	209	4	on	on	IN
3262	209	5	content	content	NN
3262	209	6	-	-	HYPH
3262	209	7	based	base	VBN
3262	209	8	music	music	NN
3262	209	9	genre	genre	NN
3262	209	10	classification	classification	NN
3262	209	11	.	.	.
3262	210	1	In	in	IN
3262	210	2	Proceedings	proceeding	NNS
3262	210	3	of	of	IN
3262	210	4	the	the	DT
3262	210	5	26th	26th	JJ
3262	210	6	Annual	Annual	NNP
3262	210	7	Interna-	Interna-	NNP
3262	210	8	tional	tional	JJ
3262	210	9	ACM	ACM	NNP
3262	210	10	SIGIR	SIGIR	NNP
3262	210	11	Conference	Conference	NNP
3262	210	12	on	on	IN
3262	210	13	Research	Research	NNP
3262	210	14	and	and	CC
3262	210	15	Development	Development	NNP
3262	210	16	in	in	IN
3262	210	17	Information	Information	NNP
3262	210	18	Retrieval	Retrieval	NNP
3262	210	19	,	,	,
3262	210	20	C.	C.	NNP
3262	210	21	Clarke	Clarke	NNP
3262	210	22	,	,	,
3262	210	23	et	et	NNP
3262	210	24	al	al	NNP
3262	210	25	.	.	NNP
3262	210	26	,	,	,
3262	210	27	eds	eds	NNP
3262	210	28	.	.	.
3262	211	1	New	New	NNP
3262	211	2	York	York	NNP
3262	211	3	:	:	:
3262	211	4	ACM	ACM	NNP
3262	211	5	.	.	.
3262	212	1	Li	Li	NNP
3262	212	2	,	,	,
3262	212	3	J.	J.	NNP
3262	212	4	and	and	CC
3262	212	5	J.	J.	NNP
3262	212	6	Wang	Wang	NNP
3262	212	7	,	,	,
3262	212	8	J.	J.	NNP
3262	213	1	2006	2006	CD
3262	213	2	.	.	.
3262	214	1	Real	real	JJ
3262	214	2	-	-	HYPH
3262	214	3	time	time	NN
3262	214	4	computerized	computerized	JJ
3262	214	5	annotation	annotation	NN
3262	214	6	of	of	IN
3262	214	7	pictures	picture	NNS
3262	214	8	.	.	.
3262	215	1	In	in	IN
3262	215	2	Proceedings	proceeding	NNS
3262	215	3	of	of	IN
3262	215	4	the	the	DT
3262	215	5	14th	14th	JJ
3262	215	6	Annual	Annual	NNP
3262	215	7	ACM	ACM	NNP
3262	215	8	International	International	NNP
3262	215	9	aRtiCLE	aRtiCLE	:
3262	215	10	titLE	title	NN
3262	215	11	|	|	NNP
3262	215	12	autHoR	autHoR	NNP
3262	215	13	47ContEnt	47ContEnt	NNP
3262	215	14	-	-	HYPH
3262	215	15	BasED	BasED	NNP
3262	215	16	inFoRmation	information	NN
3262	215	17	REtRiEvaL	REtRiEvaL	NNP
3262	215	18	anD	and	CC
3262	215	19	DiGitaL	DiGitaL	NNP
3262	215	20	LiBRaRiEs	LiBRaRiEs	NNP
3262	215	21	|	|	NNP
3262	215	22	Wan	Wan	NNP
3262	215	23	anD	and	CC
3262	215	24	Liu	Liu	NNP
3262	215	25	47	47	CD
3262	215	26	Conference	Conference	NNP
3262	215	27	on	on	IN
3262	215	28	Multimedia	Multimedia	NNP
3262	215	29	,	,	,
3262	215	30	K.	K.	NNP
3262	215	31	Nahrstedt	Nahrstedt	NNP
3262	215	32	,	,	,
3262	215	33	et	et	NNP
3262	215	34	al	al	NNP
3262	215	35	.	.	NNP
3262	215	36	,	,	,
3262	215	37	eds	eds	NNP
3262	215	38	.	.	.
3262	216	1	New	New	NNP
3262	216	2	York	York	NNP
3262	216	3	:	:	:
3262	216	4	ACM	ACM	NNP
3262	216	5	.	.	.
3262	217	1	Marques	Marques	NNP
3262	217	2	,	,	,
3262	217	3	O.	O.	NNP
3262	217	4	and	and	CC
3262	217	5	B.	B.	NNP
3262	217	6	Furht	Furht	NNP
3262	217	7	.	.	.
3262	218	1	2002	2002	CD
3262	218	2	.	.	.
3262	219	1	Content	content	NN
3262	219	2	-	-	HYPH
3262	219	3	based	base	VBN
3262	219	4	Image	Image	NNP
3262	219	5	and	and	CC
3262	219	6	Video	Video	NNP
3262	219	7	Retrieval	Retrieval	NNP
3262	219	8	.	.	.
3262	220	1	Norwell	Norwell	NNP
3262	220	2	,	,	,
3262	220	3	Mass	Mass	NNP
3262	220	4	:	:	:
3262	220	5	Kluwer	Kluwer	NNP
3262	220	6	.	.	.
3262	221	1	Moustakas	Moustakas	NNP
3262	221	2	,	,	,
3262	221	3	K.	K.	NNP
3262	221	4	,	,	,
3262	221	5	et	et	NNP
3262	221	6	al	al	NNP
3262	221	7	.	.	.
3262	222	1	2005	2005	CD
3262	222	2	.	.	.
3262	223	1	MASTER	master	NN
3262	223	2	-	-	HYPH
3262	223	3	PIECE	PIECE	NNP
3262	223	4	:	:	:
3262	223	5	A	a	DT
3262	223	6	multimodal	multimodal	NN
3262	223	7	(	(	-LRB-
3262	223	8	gesture+speech	gesture+speech	NNP
3262	223	9	)	)	-RRB-
3262	223	10	interface	interface	NN
3262	223	11	for	for	IN
3262	223	12	3D	3d	JJ
3262	223	13	model	model	NN
3262	223	14	search	search	NN
3262	223	15	and	and	CC
3262	223	16	retrieval	retrieval	NN
3262	223	17	integrated	integrate	VBN
3262	223	18	in	in	IN
3262	223	19	a	a	DT
3262	223	20	virtual	virtual	JJ
3262	223	21	assembly	assembly	NN
3262	223	22	application	application	NN
3262	223	23	.	.	.
3262	224	1	Proceedings	proceeding	NNS
3262	224	2	of	of	IN
3262	224	3	the	the	DT
3262	224	4	eNTERFACE	enterface	NN
3262	224	5	:	:	:
3262	224	6	62–75	62–75	CD
3262	224	7	.	.	.
3262	225	1	Wang	Wang	NNP
3262	225	2	,	,	,
3262	225	3	J.	J.	NNP
3262	225	4	,	,	,
3262	225	5	et	et	NNP
3262	225	6	al	al	NNP
3262	225	7	.	.	.
3262	226	1	2001	2001	CD
3262	226	2	.	.	.
3262	227	1	SIMPLIcity	simplicity	NN
3262	227	2	:	:	:
3262	227	3	semantics	semantic	NNS
3262	227	4	-	-	HYPH
3262	227	5	sensitive	sensitive	JJ
3262	227	6	integrated	integrated	JJ
3262	227	7	matching	matching	NN
3262	227	8	for	for	IN
3262	227	9	picture	picture	NN
3262	227	10	libraries	library	NNS
3262	227	11	.	.	.
3262	228	1	IEEE	IEEE	NNP
3262	228	2	Trans	Trans	NNPS
3262	228	3	.	.	.
3262	229	1	Pattern	Pattern	NNP
3262	229	2	Analysis	Analysis	NNP
3262	229	3	and	and	CC
3262	229	4	Machine	Machine	NNP
3262	229	5	Intelligence	Intelligence	NNP
3262	229	6	23	23	CD
3262	229	7	,	,	,
3262	229	8	no	no	UH
3262	229	9	.	.	.
3262	230	1	9	9	CD
3262	230	2	:	:	:
3262	230	3	947–963	947–963	CD
3262	230	4	.	.	.
3262	231	1	Wang	Wang	NNP
3262	231	2	,	,	,
3262	231	3	Z.	Z.	NNP
3262	231	4	,	,	,
3262	231	5	et	et	NNP
3262	231	6	al	al	NNP
3262	231	7	.	.	.
3262	232	1	2006	2006	CD
3262	232	2	.	.	.
3262	233	1	VFerret	vferret	NN
3262	233	2	:	:	:
3262	233	3	content	content	NN
3262	233	4	-	-	HYPH
3262	233	5	based	base	VBN
3262	233	6	similarity	similarity	NN
3262	233	7	search	search	NN
3262	233	8	tool	tool	NN
3262	233	9	for	for	IN
3262	233	10	continuous	continuous	JJ
3262	233	11	archived	archived	JJ
3262	233	12	video	video	NN
3262	233	13	.	.	.
3262	234	1	In	in	IN
3262	234	2	Proceedings	Proceedings	NNP
3262	234	3	of	of	IN
3262	234	4	the	the	DT
3262	234	5	3rd	3rd	JJ
3262	234	6	ACM	ACM	NNP
3262	234	7	Workshop	Workshop	NNP
3262	234	8	on	on	IN
3262	234	9	Continuous	Continuous	NNP
3262	234	10	Archival	Archival	NNP
3262	234	11	and	and	CC
3262	234	12	Retrival	Retrival	NNP
3262	234	13	of	of	IN
3262	234	14	Personal	Personal	NNP
3262	234	15	Experiences	Experiences	NNPS
3262	234	16	,	,	,
3262	234	17	K.	K.	NNP
3262	234	18	Maze	Maze	NNP
3262	234	19	et	et	NNP
3262	234	20	al	al	NNP
3262	234	21	.	.	NNP
3262	234	22	,	,	,
3262	234	23	eds	eds	NNP
3262	234	24	.	.	.
3262	235	1	New	New	NNP
3262	235	2	York	York	NNP
3262	235	3	:	:	:
3262	235	4	ACM	ACM	NNP
3262	235	5	.	.	.
3262	236	1	Wold	Wold	NNP
3262	236	2	,	,	,
3262	236	3	E.	E.	NNP
3262	236	4	,	,	,
3262	236	5	et	et	NNP
3262	236	6	al	al	NNP
3262	236	7	.	.	.
3262	237	1	1996	1996	CD
3262	237	2	.	.	.
3262	238	1	Content	content	NN
3262	238	2	-	-	HYPH
3262	238	3	based	base	VBN
3262	238	4	classification	classification	NN
3262	238	5	,	,	,
3262	238	6	search	search	NN
3262	238	7	,	,	,
3262	238	8	and	and	CC
3262	238	9	retrieval	retrieval	NN
3262	238	10	of	of	IN
3262	238	11	audio	audio	NNP
3262	238	12	.	.	.
3262	239	1	IEEE	IEEE	NNP
3262	239	2	MultiMedia	MultiMedia	NNP
3262	239	3	3	3	CD
3262	239	4	,	,	,
3262	239	5	no	no	UH
3262	239	6	.	.	.
3262	240	1	3	3	CD
3262	240	2	:	:	SYM
3262	240	3	27–36	27–36	CD
3262	240	4	.	.	.
3262	241	1	Zhang	Zhang	NNP
3262	241	2	,	,	,
3262	241	3	T.	T.	NNP
3262	241	4	and	and	CC
3262	241	5	C.	C.	NNP
3262	241	6	Kuo	Kuo	NNP
3262	241	7	.	.	.
3262	242	1	2001	2001	CD
3262	242	2	.	.	.
3262	243	1	Content	content	NN
3262	243	2	-	-	HYPH
3262	243	3	based	base	VBN
3262	243	4	Audio	Audio	NNP
3262	243	5	Classification	Classification	NNP
3262	243	6	and	and	CC
3262	243	7	Retrieval	Retrieval	NNP
3262	243	8	for	for	IN
3262	243	9	Audiovisual	Audiovisual	NNP
3262	243	10	Data	Data	NNP
3262	243	11	Parsing	Parsing	NNP
3262	243	12	.	.	.
3262	244	1	Norwell	Norwell	NNP
3262	244	2	,	,	,
3262	244	3	Mass.	Massachusetts	NNP
3262	245	1	:	:	:
3262	245	2	Kluwer	Kluwer	NNP
3262	245	3	.	.	.
3262	246	1	LITA	LITA	NNP
3262	246	2	National	National	NNP
3262	246	3	Forum	Forum	NNP
3262	246	4	cover	cover	NN
3262	246	5	2	2	CD
3262	246	6	LITA	LITA	NNP
3262	246	7	Guides	Guides	NNPS
3262	246	8	cover	cover	VBP
3262	246	9	3	3	CD
3262	246	10	LITA	LITA	NNP
3262	246	11	Workshops	Workshops	NNPS
3262	246	12	cover	cover	VBP
3262	246	13	4	4	CD
3262	246	14	Index	Index	NNP
3262	246	15	to	to	IN
3262	246	16	Advertisers	Advertisers	NNPS
3262	246	17	STATEMENT	statement	NN
3262	246	18	OF	of	IN
3262	246	19	OWNERSHIP	OWNERSHIP	NNP
3262	246	20	,	,	,
3262	246	21	MANAGEMENT	MANAGEMENT	NNP
3262	246	22	,	,	,
3262	246	23	AND	and	CC
3262	246	24	CIRCULATION	CIRCULATION	NNP
3262	246	25	Information	Information	NNP
3262	246	26	Technology	Technology	NNP
3262	246	27	and	and	CC
3262	246	28	Libraries	Libraries	NNPS
3262	246	29	,	,	,
3262	246	30	Publication	Publication	NNP
3262	246	31	No	no	NN
3262	246	32	.	.	.
3262	247	1	280	280	CD
3262	247	2	-	-	SYM
3262	247	3	800	800	CD
3262	247	4	,	,	,
3262	247	5	is	be	VBZ
3262	247	6	published	publish	VBN
3262	247	7	quarterly	quarterly	RB
3262	247	8	in	in	IN
3262	247	9	March	March	NNP
3262	247	10	,	,	,
3262	247	11	June	June	NNP
3262	247	12	,	,	,
3262	247	13	September	September	NNP
3262	247	14	,	,	,
3262	247	15	and	and	CC
3262	247	16	December	December	NNP
3262	247	17	by	by	IN
3262	247	18	the	the	DT
3262	247	19	Library	Library	NNP
3262	247	20	Information	Information	NNP
3262	247	21	and	and	CC
3262	247	22	Technology	Technology	NNP
3262	247	23	Association	Association	NNP
3262	247	24	,	,	,
3262	247	25	American	American	NNP
3262	247	26	Library	Library	NNP
3262	247	27	Association	Association	NNP
3262	247	28	,	,	,
3262	247	29	50	50	CD
3262	247	30	E.	E.	NNP
3262	247	31	Huron	Huron	NNP
3262	247	32	St.	St.	NNP
3262	247	33	,	,	,
3262	247	34	Chicago	Chicago	NNP
3262	247	35	,	,	,
3262	247	36	Illinois	Illinois	NNP
3262	247	37	60611	60611	CD
3262	247	38	-	-	SYM
3262	247	39	2795	2795	CD
3262	247	40	.	.	.
3262	248	1	Editor	editor	NN
3262	248	2	:	:	:
3262	248	3	John	John	NNP
3262	248	4	Webb	Webb	NNP
3262	248	5	,	,	,
3262	248	6	Librarian	Librarian	NNP
3262	248	7	Emeritus	Emeritus	NNP
3262	248	8	,	,	,
3262	248	9	Washington	Washington	NNP
3262	248	10	State	State	NNP
3262	248	11	University	University	NNP
3262	248	12	Libraries	Libraries	NNPS
3262	248	13	,	,	,
3262	248	14	Pullman	Pullman	NNP
3262	248	15	,	,	,
3262	248	16	WA	WA	NNP
3262	248	17	99164	99164	CD
3262	248	18	-	-	HYPH
3262	248	19	5610	5610	CD
3262	248	20	.	.	.
3262	249	1	Annual	annual	JJ
3262	249	2	subscription	subscription	NN
3262	249	3	price	price	NN
3262	249	4	,	,	,
3262	249	5	$	$	$
3262	249	6	55	55	CD
3262	249	7	.	.	.
3262	250	1	Printed	Printed	NNP
3262	250	2	in	in	IN
3262	250	3	U.S.A.	U.S.A.	NNP
3262	250	4	with	with	IN
3262	250	5	periodical	periodical	JJ
3262	250	6	-	-	HYPH
3262	250	7	class	class	NN
3262	250	8	postage	postage	NN
3262	250	9	paid	pay	VBN
3262	250	10	at	at	IN
3262	250	11	Chicago	Chicago	NNP
3262	250	12	,	,	,
3262	250	13	Illinois	Illinois	NNP
3262	250	14	,	,	,
3262	250	15	and	and	CC
3262	250	16	other	other	JJ
3262	250	17	locations	location	NNS
3262	250	18	.	.	.
3262	251	1	As	as	IN
3262	251	2	a	a	DT
3262	251	3	nonprofit	nonprofit	JJ
3262	251	4	organization	organization	NN
3262	251	5	authorized	authorize	VBN
3262	251	6	to	to	TO
3262	251	7	mail	mail	VB
3262	251	8	at	at	IN
3262	251	9	special	special	JJ
3262	251	10	rates	rate	NNS
3262	251	11	(	(	-LRB-
3262	251	12	DMM	DMM	NNP
3262	251	13	Section	Section	NNP
3262	251	14	424.12	424.12	CD
3262	251	15	only	only	RB
3262	251	16	)	)	-RRB-
3262	251	17	,	,	,
3262	251	18	the	the	DT
3262	251	19	purpose	purpose	NN
3262	251	20	,	,	,
3262	251	21	function	function	NN
3262	251	22	,	,	,
3262	251	23	and	and	CC
3262	251	24	nonprofit	nonprofit	JJ
3262	251	25	status	status	NN
3262	251	26	for	for	IN
3262	251	27	federal	federal	JJ
3262	251	28	income	income	NN
3262	251	29	tax	tax	NN
3262	251	30	purposes	purpose	NNS
3262	251	31	have	have	VBP
3262	251	32	not	not	RB
3262	251	33	changed	change	VBN
3262	251	34	during	during	IN
3262	251	35	the	the	DT
3262	251	36	preceding	precede	VBG
3262	251	37	twelve	twelve	CD
3262	251	38	months	month	NNS
3262	251	39	.	.	.
3262	252	1	EXTENT	extent	NN
3262	252	2	AND	and	CC
3262	252	3	NATURE	nature	NN
3262	252	4	OF	of	IN
3262	252	5	CIRCULATION	circulation	NN
3262	252	6	(	(	-LRB-
3262	252	7	Average	average	JJ
3262	252	8	figures	figure	NNS
3262	252	9	denote	denote	VBP
3262	252	10	the	the	DT
3262	252	11	average	average	JJ
3262	252	12	number	number	NN
3262	252	13	of	of	IN
3262	252	14	copies	copy	NNS
3262	252	15	printed	print	VBN
3262	252	16	each	each	DT
3262	252	17	issue	issue	NN
3262	252	18	during	during	IN
3262	252	19	the	the	DT
3262	252	20	preceding	precede	VBG
3262	252	21	twelve	twelve	CD
3262	252	22	months	month	NNS
3262	252	23	;	;	:
3262	252	24	actual	actual	JJ
3262	252	25	figures	figure	NNS
3262	252	26	denote	denote	VBP
3262	252	27	actual	actual	JJ
3262	252	28	number	number	NN
3262	252	29	of	of	IN
3262	252	30	copies	copy	NNS
3262	252	31	of	of	IN
3262	252	32	single	single	JJ
3262	252	33	issue	issue	NN
3262	252	34	published	publish	VBD
3262	252	35	nearest	near	JJS
3262	252	36	to	to	IN
3262	252	37	filing	filing	NN
3262	252	38	date	date	NN
3262	252	39	:	:	:
3262	252	40	June	June	NNP
3262	252	41	2007	2007	CD
3262	252	42	issue	issue	NN
3262	252	43	)	)	-RRB-
3262	252	44	.	.	.
3262	253	1	Total	total	JJ
3262	253	2	number	number	NN
3262	253	3	of	of	IN
3262	253	4	copies	copy	NNS
3262	253	5	printed	print	VBN
3262	253	6	:	:	:
3262	253	7	average	average	NNP
3262	253	8	,	,	,
3262	253	9	5,354	5,354	CD
3262	253	10	;	;	:
3262	253	11	actual	actual	JJ
3262	253	12	,	,	,
3262	253	13	5,280	5,280	CD
3262	253	14	.	.	.
3262	254	1	Sales	sale	NNS
3262	254	2	through	through	IN
3262	254	3	dealers	dealer	NNS
3262	254	4	and	and	CC
3262	254	5	carriers	carrier	NNS
3262	254	6	,	,	,
3262	254	7	street	street	NN
3262	254	8	vendors	vendor	NNS
3262	254	9	,	,	,
3262	254	10	and	and	CC
3262	254	11	counter	counter	NN
3262	254	12	sales	sale	NNS
3262	254	13	:	:	:
3262	254	14	average	average	JJ
3262	254	15	,	,	,
3262	254	16	0	0	CD
3262	254	17	;	;	:
3262	254	18	actual	actual	JJ
3262	254	19	462	462	CD
3262	254	20	.	.	.
3262	255	1	Paid	pay	VBN
3262	255	2	or	or	CC
3262	255	3	requested	request	VBN
3262	255	4	mail	mail	NN
3262	255	5	subscriptions	subscription	NNS
3262	255	6	:	:	:
3262	255	7	average	average	JJ
3262	255	8	,	,	,
3262	255	9	4,283	4,283	CD
3262	255	10	;	;	:
3262	255	11	actual	actual	JJ
3262	255	12	,	,	,
3262	255	13	4,193	4,193	CD
3262	255	14	.	.	.
3262	256	1	Free	free	JJ
3262	256	2	distribution	distribution	NN
3262	256	3	(	(	-LRB-
3262	256	4	total	total	NN
3262	256	5	)	)	-RRB-
3262	256	6	:	:	:
3262	256	7	average	average	JJ
3262	256	8	,	,	,
3262	256	9	292	292	CD
3262	256	10	;	;	:
3262	256	11	actual	actual	JJ
3262	256	12	,	,	,
3262	256	13	292	292	CD
3262	256	14	.	.	.
3262	257	1	Total	total	JJ
3262	257	2	distribution	distribution	NN
3262	257	3	:	:	:
3262	257	4	average	average	JJ
3262	257	5	,	,	,
3262	257	6	5,028	5,028	CD
3262	257	7	;	;	:
3262	257	8	actual	actual	JJ
3262	257	9	,	,	,
3262	257	10	4,947	4,947	CD
3262	257	11	.	.	.
3262	258	1	Office	office	NN
3262	258	2	use	use	NN
3262	258	3	,	,	,
3262	258	4	leftover	leftover	NN
3262	258	5	,	,	,
3262	258	6	unaccounted	unaccounted	JJ
3262	258	7	,	,	,
3262	258	8	spoiled	spoil	VBD
3262	258	9	after	after	IN
3262	258	10	printing	print	VBG
3262	258	11	:	:	:
3262	258	12	average	average	JJ
3262	258	13	,	,	,
3262	258	14	326	326	CD
3262	258	15	;	;	:
3262	258	16	actual	actual	JJ
3262	258	17	,	,	,
3262	258	18	333	333	CD
3262	258	19	.	.	.
3262	259	1	Total	total	NN
3262	259	2	:	:	:
3262	259	3	average	average	JJ
3262	259	4	,	,	,
3262	259	5	5,354	5,354	CD
3262	259	6	;	;	:
3262	259	7	actual	actual	JJ
3262	259	8	,	,	,
3262	259	9	5,280	5,280	CD
3262	259	10	.	.	.
3262	260	1	Percentage	percentage	NN
3262	260	2	paid	pay	VBN
3262	260	3	:	:	:
3262	260	4	average	average	JJ
3262	260	5	,	,	,
3262	260	6	94.19	94.19	CD
3262	260	7	;	;	:
3262	260	8	actual	actual	JJ
3262	260	9	,	,	,
3262	260	10	94.10	94.10	CD
3262	260	11	.	.	.
3262	261	1	S	S	NNP
3262	261	2	t	t	NNP
3262	261	3	a	a	DT
3262	261	4	t	t	NN
3262	261	5	e	e	NNP
3262	261	6	m	m	NNP
3262	261	7	e	e	NNP
3262	261	8	n	n	NNP
3262	261	9	t	t	NNP
3262	261	10	o	o	NN
3262	261	11	f	f	NN
3262	261	12	O	o	NN
3262	261	13	w	w	NN
3262	261	14	n	n	NNP
3262	261	15	e	e	NNP
3262	261	16	r	r	NNP
3262	261	17	s	s	NNP
3262	261	18	h	h	NNP
3262	261	19	i	i	NNP
3262	261	20	p	p	NN
3262	261	21	,	,	,
3262	261	22	M	M	NNP
3262	261	23	a	a	DT
3262	261	24	n	n	NN
3262	261	25	a	a	NN
3262	261	26	g	g	NN
3262	261	27	e	e	NNP
3262	261	28	m	m	NNP
3262	261	29	e	e	NNP
3262	261	30	n	n	NNP
3262	261	31	t	t	NNP
3262	261	32	,	,	,
3262	261	33	a	a	DT
3262	261	34	n	n	NNP
3262	261	35	d	d	NNP
3262	261	36	C	c	NN
3262	261	37	i	i	NN
3262	261	38	r	r	NN
3262	261	39	c	c	NN
3262	261	40	u	u	NNP
3262	261	41	l	l	NNP
3262	261	42	a	a	DT
3262	261	43	t	t	NN
3262	261	44	i	i	NN
3262	261	45	o	o	XX
3262	261	46	n	n	NN
3262	261	47	(	(	-LRB-
3262	261	48	P	p	NN
3262	261	49	S	S	NNP
3262	261	50	F	F	NNP
3262	261	51	o	o	NN
3262	261	52	r	r	NNP
3262	261	53	m	m	NNP
3262	261	54	3	3	CD
3262	261	55	5	5	CD
3262	261	56	2	2	CD
3262	261	57	6	6	CD
3262	261	58	,	,	,
3262	261	59	S	S	NNP
3262	261	60	e	e	NNP
3262	261	61	p	p	NN
3262	261	62	t	t	NNP
3262	261	63	e	e	NNP
3262	261	64	m	m	NNP
3262	261	65	b	b	NNP
3262	261	66	e	e	NNP
3262	261	67	r	r	NNP
3262	261	68	2	2	CD
3262	261	69	0	0	CD
3262	261	70	0	0	CD
3262	261	71	7	7	CD
3262	261	72	)	)	-RRB-
3262	261	73	f	f	NNP
3262	261	74	i	i	NN
3262	261	75	l	l	NNP
3262	261	76	e	e	NNP
3262	261	77	d	d	NNP
3262	261	78	w	w	NNP
3262	261	79	i	i	NNP
3262	261	80	t	t	NNP
3262	261	81	h	h	NNP
3262	261	82	t	t	NN
3262	261	83	h	h	NN
3262	261	84	e	e	NN
3262	261	85	U	u	NN
3262	261	86	n	n	NN
3262	261	87	i	i	NN
3262	261	88	t	t	NN
3262	261	89	e	e	NNP
3262	261	90	d	d	NNP
3262	261	91	S	S	NNP
3262	261	92	t	t	NN
3262	261	93	a	a	DT
3262	261	94	t	t	NN
3262	261	95	e	e	NN
3262	261	96	s	s	NNP
3262	261	97	P	p	NN
3262	261	98	o	o	NN
3262	261	99	s	s	NN
3262	261	100	t	t	NN
3262	261	101	O	o	NN
3262	261	102	f	f	NNP
3262	261	103	f	f	NN
3262	261	104	i	i	NN
3262	261	105	c	c	NN
3262	261	106	e	e	NNP
3262	261	107	P	p	NN
3262	261	108	o	o	NN
3262	261	109	s	s	NNP
3262	261	110	t	t	NNP
3262	261	111	m	m	NNP
3262	261	112	a	a	NNP
3262	261	113	s	s	NNP
3262	261	114	t	t	NNP
3262	261	115	e	e	NNP
3262	261	116	r	r	NN
3262	261	117	i	i	NN
3262	261	118	n	n	CC
3262	261	119	C	C	NNP
3262	261	120	h	h	NN
3262	261	121	i	i	NN
3262	261	122	c	c	NN
3262	261	123	a	a	DT
3262	261	124	g	g	NN
3262	261	125	o	o	NN
3262	261	126	,	,	,
3262	261	127	O	o	UH
3262	261	128	c	c	NN
3262	261	129	t	t	NN
3262	261	130	o	o	NNP
3262	261	131	b	b	NNP
3262	261	132	e	e	NNP
3262	261	133	r	r	NNP
3262	261	134	1	1	CD
3262	261	135	,	,	,
3262	261	136	2	2	CD
3262	261	137	0	0	CD
3262	261	138	0	0	CD
3262	261	139	7	7	CD
3262	261	140	.	.	.