id	sid	tid	token	lemma	pos
work_nvbfnacgujehpi52fk2nc7sg2m	1	1	OP	OP	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	2	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	1	3	LLCJ140051	LLCJ140051	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	4	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	1	5	..	..	NFP
work_nvbfnacgujehpi52fk2nc7sg2m	1	6	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	1	7	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	1	8	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	10	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	11	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	12	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	1	13	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	14	digital	digital	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	15	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	16	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	17	............................................................................................................................................................	............................................................................................................................................................	.
work_nvbfnacgujehpi52fk2nc7sg2m	1	18	Ofer	Ofer	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	19	Biller	Biller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	1	21	Jihad	Jihad	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	22	El	El	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	23	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	1	24	Sana	Sana	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	25	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	1	26	Klara	Klara	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	27	Kedem	Kedem	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	28	Ben	Ben	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	29	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	1	30	Gurion	Gurion	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	31	University	University	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	32	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	33	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	1	34	Negev	Negev	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	35	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	1	36	Israel	Israel	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	37	.......................................................................................................................................	.......................................................................................................................................	.
work_nvbfnacgujehpi52fk2nc7sg2m	1	38	Abstract	Abstract	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	1	39	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	1	40	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	41	studies	study	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	1	42	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	1	43	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	44	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	45	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	46	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	47	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	1	48	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	49	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	1	50	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	51	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	52	digital	digital	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	53	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	54	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	1	55	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	56	degraded	degraded	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	57	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	1	58	such	such	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	59	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	1	60	histor-	histor-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	61	ical	ical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	1	62	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	1	63	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	2	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	2	2	provide	provide	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	2	3	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	2	4	rigorous	rigorous	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	2	5	scheme	scheme	NN
work_nvbfnacgujehpi52fk2nc7sg2m	2	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	2	7	quantifying	quantify	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	2	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	2	9	statistical	statistical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	2	10	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	2	11	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	2	12	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	2	13	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	2	14	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	2	15	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	2	16	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	2	17	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	2	18	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	2	19	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	2	20	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	2	21	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	2	22	such	such	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	2	23	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	2	24	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	3	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	3	2	study	study	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	3	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	3	4	compare	compare	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	3	5	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	3	6	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	3	7	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	3	8	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	3	9	four	four	CD
work_nvbfnacgujehpi52fk2nc7sg2m	3	10	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	3	11	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	3	12	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	3	13	measure	measure	VB
work_nvbfnacgujehpi52fk2nc7sg2m	3	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	3	15	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	3	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	3	17	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	3	18	individual	individual	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	3	19	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	3	20	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	3	21	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	3	22	digital	digital	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	3	23	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	3	24	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	3	25	process	process	NN
work_nvbfnacgujehpi52fk2nc7sg2m	3	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	4	1	To	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	4	2	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	4	3	end	end	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	4	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	4	6	create	create	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	4	7	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	4	8	lan-	lan-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	9	guages	guage	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	4	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	4	11	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	4	12	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	4	13	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	4	14	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	4	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	4	16	except	except	IN
work_nvbfnacgujehpi52fk2nc7sg2m	4	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	4	18	one	one	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	19	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	4	20	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	4	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	4	22	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	4	23	identical	identical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	4	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	4	25	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	4	26	measure	measure	VB
work_nvbfnacgujehpi52fk2nc7sg2m	4	27	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	4	28	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	29	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	4	30	two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	4	31	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	32	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	33	algorithms	algorithm	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	4	34	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	4	35	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	4	36	docu-	docu-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	4	37	ments	ment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	4	38	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	4	39	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	4	40	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	4	41	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	5	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	5	2	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	5	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	5	4	summarize	summarize	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	5	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	5	6	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	5	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	5	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	5	9	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	5	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	5	11	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	5	12	characteristic	characteristic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	5	13	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	5	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	5	15	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	5	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	5	17	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	5	18	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	5	19	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	5	20	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	5	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	6	1	.................................................................................................................................................................................	.................................................................................................................................................................................	NFP
work_nvbfnacgujehpi52fk2nc7sg2m	6	2	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	6	3	Introduction	Introduction	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	6	4	Research	Research	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	6	5	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	6	6	digital	digital	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	6	7	script	script	NN
work_nvbfnacgujehpi52fk2nc7sg2m	6	8	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	6	9	could	could	MD
work_nvbfnacgujehpi52fk2nc7sg2m	6	10	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	6	11	clas-	clas-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	6	12	sified	sifie	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	6	13	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	6	14	two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	6	15	main	main	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	6	16	approaches	approach	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	6	17	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	6	18	segmentation	segmentation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	6	19	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	6	20	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	6	21	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	6	22	segmentation	segmentation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	6	23	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	6	24	free	free	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	6	25	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	6	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	7	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	7	2	segmenta-	segmenta-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	7	3	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	7	4	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	7	5	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	7	6	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	7	7	segments	segment	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	7	8	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	7	9	input	input	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	7	10	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	7	11	into	into	IN
work_nvbfnacgujehpi52fk2nc7sg2m	7	12	individual	individual	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	7	13	characters	character	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	7	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	7	15	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	7	16	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	7	17	then	then	RB
work_nvbfnacgujehpi52fk2nc7sg2m	7	18	recognized	recognize	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	7	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	7	20	combined	combine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	7	21	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	7	22	identify	identify	VB
work_nvbfnacgujehpi52fk2nc7sg2m	7	23	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	7	24	input	input	NN
work_nvbfnacgujehpi52fk2nc7sg2m	7	25	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	7	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	8	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	8	2	seg-	seg-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	8	3	mentation	mentation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	8	4	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	8	5	free	free	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	8	6	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	8	7	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	8	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	8	9	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	8	10	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	8	11	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	8	12	recognizes	recognize	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	8	13	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	8	14	whole	whole	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	8	15	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	8	16	without	without	IN
work_nvbfnacgujehpi52fk2nc7sg2m	8	17	segmenting	segment	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	8	18	it	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	8	19	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	8	20	opposed	oppose	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	8	21	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	8	22	classical	classical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	8	23	Optical	Optical	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	8	24	Character	Character	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	8	25	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	8	26	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	8	27	OCR	OCR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	8	28	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	8	29	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	8	30	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	9	1	Recently	recently	RB
work_nvbfnacgujehpi52fk2nc7sg2m	9	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	9	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	9	4	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	9	5	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	9	7	digital	digital	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	9	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	9	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	10	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	9	11	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	9	12	attracting	attract	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	9	13	more	more	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	9	14	interest	interest	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	9	16	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	9	17	become	become	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	9	18	widely	widely	RB
work_nvbfnacgujehpi52fk2nc7sg2m	9	19	accepted	accept	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	9	20	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	9	21	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	9	22	handwriting	handwriting	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	23	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	24	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	25	community	community	NN
work_nvbfnacgujehpi52fk2nc7sg2m	9	26	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	9	27	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	10	1	Plamondon	Plamondon	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	10	3	Guerfali	Guerfali	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	4	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	10	5	1996	1996	CD
work_nvbfnacgujehpi52fk2nc7sg2m	10	6	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	10	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	10	8	Plamondon	Plamondon	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	10	10	Srihari	Srihari	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	11	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	10	12	2000	2000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	10	13	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	10	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	10	15	Gatos	Gatos	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	16	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	17	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	10	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	11	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	11	2	2005	2005	CD
work_nvbfnacgujehpi52fk2nc7sg2m	11	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	11	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	11	5	Biadsy	Biadsy	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	11	6	et	et	FW
work_nvbfnacgujehpi52fk2nc7sg2m	11	7	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	11	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	12	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	12	2	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	12	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	12	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	12	5	Zagoris	Zagoris	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	12	6	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	12	7	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	12	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	13	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	13	2	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	13	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	13	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	14	1	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	14	2	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	3	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	14	4	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	14	5	adapted	adapt	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	14	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	14	7	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	8	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	14	10	historical	historical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	14	11	docu-	docu-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	12	ments	ment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	14	13	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	14	14	usually	usually	RB
work_nvbfnacgujehpi52fk2nc7sg2m	14	15	suffer	suffer	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	14	16	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	14	17	high	high	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	14	18	level	level	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	19	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	14	20	deg-	deg-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	21	radation	radation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	14	22	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	14	23	Lavrenko	Lavrenko	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	14	24	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	14	25	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	14	26	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	14	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	14	28	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	14	29	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	14	30	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	14	31	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	14	32	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	14	33	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	14	34	2007	2007	CD
work_nvbfnacgujehpi52fk2nc7sg2m	14	35	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	14	36	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	15	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	15	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	15	3	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	15	4	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	15	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	15	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	15	7	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	15	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	15	9	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	15	10	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	15	11	depends	depend	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	15	12	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	15	13	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	15	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	15	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	15	16	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	15	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	16	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	16	2	example	example	NN
work_nvbfnacgujehpi52fk2nc7sg2m	16	3	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	16	4	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	16	5	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	16	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	16	7	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	16	8	there	there	EX
work_nvbfnacgujehpi52fk2nc7sg2m	16	9	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	16	10	many	many	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	16	11	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	16	12	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	16	13	similar	similar	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	16	14	characters	character	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	16	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	16	16	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	16	17	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	16	18	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	16	19	lexicon	lexicon	NN
work_nvbfnacgujehpi52fk2nc7sg2m	16	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	16	21	may	may	MD
work_nvbfnacgujehpi52fk2nc7sg2m	16	22	have	have	VB
work_nvbfnacgujehpi52fk2nc7sg2m	16	23	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	16	24	higher	high	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	16	25	chance	chance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	16	26	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	16	27	being	be	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	16	28	misclassified	misclassifie	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	16	29	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	17	1	Such	such	PDT
work_nvbfnacgujehpi52fk2nc7sg2m	17	2	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	17	3	observation	observation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	17	4	calls	call	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	17	5	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	17	6	examining	examine	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	17	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	17	8	relation	relation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	17	9	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	17	10	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	17	11	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	17	12	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	17	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	17	14	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	17	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	17	16	automatic	automatic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	17	17	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	17	18	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	17	19	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	18	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	18	2	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	18	4	sev-	sev-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	18	5	eral	eral	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	18	6	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	7	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	18	8	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	18	9	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	18	10	visual	visual	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	18	11	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	12	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	13	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	18	14	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	18	15	studied	study	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	18	16	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	18	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	18	18	domain	domain	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	19	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	18	20	cognitive	cognitive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	18	21	psychology	psychology	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	22	perception	perception	NN
work_nvbfnacgujehpi52fk2nc7sg2m	18	23	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	18	24	Richards	Richards	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	25	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	18	26	Heller	Heller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	28	1976	1976	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	29	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	18	30	Coltheart	Coltheart	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	31	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	32	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	33	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	34	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	35	1977	1977	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	36	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	18	37	Andrews	Andrews	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	38	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	39	1989	1989	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	40	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	41	1992	1992	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	42	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	43	1997	1997	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	44	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	18	45	Grainger	Grainger	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	46	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	47	1990	1990	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	48	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	18	49	Perea	Perea	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	50	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	18	51	Rosa	Rosa	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	52	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	53	2000	2000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	54	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	18	55	New	New	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	56	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	57	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	58	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	18	59	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	18	60	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	18	61	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	18	62	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	19	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	19	2	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	19	3	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	19	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	19	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	19	6	explore	explore	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	19	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	19	8	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	19	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	19	10	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	19	11	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	19	12	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	19	13	automatic	automatic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	19	14	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	19	15	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	19	16	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	19	17	low	low	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	19	18	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	19	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	19	20	degraded	degraded	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	19	21	docu-	docu-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	19	22	ments	ment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	19	23	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	19	24	such	such	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	19	25	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	19	26	historical	historical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	19	27	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	19	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	20	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	20	2	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	20	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	20	4	following	follow	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	20	5	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	20	6	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	20	7	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	20	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	20	9	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	20	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	20	11	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	20	12	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	20	13	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	20	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	20	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	20	16	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	20	17	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	20	18	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	20	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	20	20	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	20	21	among	among	IN
work_nvbfnacgujehpi52fk2nc7sg2m	20	22	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	20	23	characters	character	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	20	24	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	21	1	Intuitively	intuitively	RB
work_nvbfnacgujehpi52fk2nc7sg2m	21	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	21	3	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	21	4	Correspondence	Correspondence	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	5	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	21	6	Ofer	Ofer	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	7	Biller	Biller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	21	9	Department	Department	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	21	11	Computer	Computer	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	12	Science	Science	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	21	14	Ben	Ben	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	15	Gurion	Gurion	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	16	University	University	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	17	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	21	18	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	21	19	Negev	Negev	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	21	21	P.O.B	P.O.B	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	22	653	653	CD
work_nvbfnacgujehpi52fk2nc7sg2m	21	23	Be’er	be’er	NN
work_nvbfnacgujehpi52fk2nc7sg2m	21	24	Sheva	Sheva	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	25	84105	84105	CD
work_nvbfnacgujehpi52fk2nc7sg2m	21	26	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	21	27	Israel	Israel	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	28	E	E	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	29	-	-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	21	30	mail	mail	NN
work_nvbfnacgujehpi52fk2nc7sg2m	21	31	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	21	32	billero@cs.bgu.ac.il	billero@cs.bgu.ac.il	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	33	Literary	literary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	21	34	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	21	35	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	36	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	21	37	�	�	.
work_nvbfnacgujehpi52fk2nc7sg2m	21	38	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	21	39	Author	author	NN
work_nvbfnacgujehpi52fk2nc7sg2m	21	40	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	21	41	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	22	1	Published	publish	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	22	2	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	22	3	Oxford	Oxford	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	22	4	University	University	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	22	5	Press	Press	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	22	6	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	22	7	behalf	behalf	NN
work_nvbfnacgujehpi52fk2nc7sg2m	22	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	22	9	EADH	EADH	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	22	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	23	1	All	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	23	2	rights	right	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	23	3	reserved	reserve	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	23	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	24	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	24	2	Permissions	permission	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	24	3	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	24	4	please	please	UH
work_nvbfnacgujehpi52fk2nc7sg2m	24	5	email	email	VB
work_nvbfnacgujehpi52fk2nc7sg2m	24	6	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	24	7	journals.permissions@oup.com	journals.permissions@oup.com	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	8	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	24	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	24	10	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	24	11	doi:10.1093	doi:10.1093	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	12	/	/	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	24	13	llc	llc	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	14	/	/	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	24	15	fqu051	fqu051	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	16	Literary	literary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	24	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	24	18	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	19	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	20	Advance	Advance	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	21	Access	Access	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	22	published	publish	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	24	23	October	October	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	24	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	24	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	24	26	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	24	27	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	24	28	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	29	en	en	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	30	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	31	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	32	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	33	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	34	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	24	35	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	36	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	37	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	38	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	39	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	24	40	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	41	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	42	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	43	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	24	44	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	24	45	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	24	46	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	47	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	48	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	49	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	24	50	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	24	51	Gatos	Gatos	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	24	52	etal	etal	NN
work_nvbfnacgujehpi52fk2nc7sg2m	24	53	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	25	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	25	2	2005	2005	CD
work_nvbfnacgujehpi52fk2nc7sg2m	25	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	25	4	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	25	5	Grainger	Grainger	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	25	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	25	7	1990	1990	CD
work_nvbfnacgujehpi52fk2nc7sg2m	25	8	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	25	9	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	25	10	statistically	statistically	RB
work_nvbfnacgujehpi52fk2nc7sg2m	25	11	influence	influence	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	25	12	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	25	13	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	25	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	25	15	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	25	16	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	25	17	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	25	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	26	1	Here	here	RB
work_nvbfnacgujehpi52fk2nc7sg2m	26	2	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	26	3	provide	provide	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	26	4	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	26	5	rigorous	rigorous	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	26	6	scheme	scheme	NN
work_nvbfnacgujehpi52fk2nc7sg2m	26	7	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	26	8	measuring	measure	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	26	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	26	10	validating	validate	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	26	11	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	26	12	intuition	intuition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	26	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	27	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	27	2	start	start	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	27	3	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	4	learning	learn	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	27	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	27	6	behavior	behavior	NN
work_nvbfnacgujehpi52fk2nc7sg2m	27	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	8	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	27	9	char-	char-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	27	10	acteristics	acteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	27	11	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	12	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	27	13	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	27	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	27	15	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	16	gathering	gather	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	27	17	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	27	18	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	27	19	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	20	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	27	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	27	22	Hebrew	Hebrew	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	27	23	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	27	24	Arabic	Arabic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	27	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	27	26	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	27	27	Russian	russian	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	27	28	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	27	29	analyzing	analyze	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	27	30	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	27	31	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	27	32	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	33	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	27	34	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	27	35	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	27	36	them	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	27	37	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	28	1	To	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	28	2	isolate	isolate	VB
work_nvbfnacgujehpi52fk2nc7sg2m	28	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	28	4	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	28	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	28	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	28	7	specific	specific	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	28	8	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	28	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	28	10	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	28	11	generate	generate	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	28	12	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	28	13	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	28	14	where	where	WRB
work_nvbfnacgujehpi52fk2nc7sg2m	28	15	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	28	16	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	28	17	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	28	18	identical	identical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	28	19	except	except	IN
work_nvbfnacgujehpi52fk2nc7sg2m	28	20	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	28	21	one	one	NN
work_nvbfnacgujehpi52fk2nc7sg2m	28	22	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	28	23	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	28	24	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	29	1	Then	then	RB
work_nvbfnacgujehpi52fk2nc7sg2m	29	2	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	29	3	produce	produce	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	29	4	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	29	5	images	image	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	29	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	29	7	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	29	8	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	29	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	29	10	apply	apply	VB
work_nvbfnacgujehpi52fk2nc7sg2m	29	11	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	29	12	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	29	13	algorithms	algorithm	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	29	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	29	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	29	16	analyze	analyze	VB
work_nvbfnacgujehpi52fk2nc7sg2m	29	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	29	18	relations	relation	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	29	19	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	29	20	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	29	21	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	29	22	characteristic	characteristic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	29	23	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	29	24	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	29	25	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	29	26	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	29	27	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	29	28	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	29	29	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	30	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	30	2	simu-	simu-	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	30	3	late	late	RB
work_nvbfnacgujehpi52fk2nc7sg2m	30	4	degraded	degraded	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	30	5	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	30	6	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	30	7	adding	add	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	30	8	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	30	9	noise	noise	NN
work_nvbfnacgujehpi52fk2nc7sg2m	30	10	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	30	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	30	12	synthetically	synthetically	RB
work_nvbfnacgujehpi52fk2nc7sg2m	30	13	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	30	14	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	30	15	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	30	16	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	30	17	con-	con-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	30	18	trolled	trolled	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	30	19	manner	manner	NN
work_nvbfnacgujehpi52fk2nc7sg2m	30	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	31	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	31	2	analysis	analysis	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	31	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	31	5	relationship	relationship	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	6	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	31	7	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	8	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	31	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	31	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	31	11	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	31	13	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	14	algorithms	algorithm	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	31	15	can	can	MD
work_nvbfnacgujehpi52fk2nc7sg2m	31	16	determine	determine	VB
work_nvbfnacgujehpi52fk2nc7sg2m	31	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	31	18	suitable	suitable	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	31	19	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	20	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	21	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	31	22	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	31	23	given	give	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	31	24	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	31	26	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	31	27	simplify	simplify	VB
work_nvbfnacgujehpi52fk2nc7sg2m	31	28	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	31	29	util-	util-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	31	30	ization	ization	NN
work_nvbfnacgujehpi52fk2nc7sg2m	31	31	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	31	32	existing	exist	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	31	33	techniques	technique	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	31	34	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	32	1	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	32	2	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	32	3	useful	useful	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	32	4	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	32	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	32	6	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	32	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	32	8	degraded	degrade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	32	9	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	32	10	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	32	11	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	32	12	lan-	lan-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	32	13	guages	guage	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	32	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	33	1	Here	here	RB
work_nvbfnacgujehpi52fk2nc7sg2m	33	2	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	33	3	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	33	4	two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	33	5	approaches	approach	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	33	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	33	7	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	33	8	spotting	spotting	NN
work_nvbfnacgujehpi52fk2nc7sg2m	33	9	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	10	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	33	11	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	12	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	33	13	dynamic	dynamic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	33	14	time	time	NN
work_nvbfnacgujehpi52fk2nc7sg2m	33	15	warping	warping	NN
work_nvbfnacgujehpi52fk2nc7sg2m	33	16	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	17	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	18	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	19	offered	offer	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	33	20	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	33	21	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	22	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	33	23	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	24	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	25	2003b	2003b	CD
work_nvbfnacgujehpi52fk2nc7sg2m	33	26	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	27	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	33	28	gradient	gradient	NN
work_nvbfnacgujehpi52fk2nc7sg2m	33	29	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	33	30	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	33	31	binary	binary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	33	32	features	feature	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	33	33	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	34	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	35	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	33	36	used	use	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	33	37	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	33	38	Zhang	Zhang	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	39	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	40	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	33	41	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	34	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	34	2	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	34	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	34	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	35	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	35	2	article	article	NN
work_nvbfnacgujehpi52fk2nc7sg2m	35	3	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	35	4	organized	organize	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	35	5	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	35	6	follows	follow	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	35	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	36	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	36	2	Section	section	NN
work_nvbfnacgujehpi52fk2nc7sg2m	36	3	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	36	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	36	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	36	6	outline	outline	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	36	7	related	related	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	36	8	work	work	NN
work_nvbfnacgujehpi52fk2nc7sg2m	36	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	37	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	37	2	Section	section	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	3	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	37	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	37	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	37	6	present	present	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	37	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	37	8	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	9	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	37	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	37	11	review	review	VB
work_nvbfnacgujehpi52fk2nc7sg2m	37	12	their	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	37	13	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	14	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	37	15	human	human	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	37	16	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	17	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	18	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	37	19	reported	report	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	37	20	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	37	21	works	work	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	37	22	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	37	23	psychology	psychology	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	24	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	37	25	cognitive	cognitive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	37	26	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	37	27	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	38	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	38	2	Section	section	NN
work_nvbfnacgujehpi52fk2nc7sg2m	38	3	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	38	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	38	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	38	6	show	show	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	38	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	38	8	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	38	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	38	10	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	38	11	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	38	12	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	38	13	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	38	14	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	38	15	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	38	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	39	1	Section	section	NN
work_nvbfnacgujehpi52fk2nc7sg2m	39	2	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	39	3	measures	measure	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	39	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	39	5	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	39	6	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	39	7	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	39	8	characteristic	characteristic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	39	9	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	39	10	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	39	11	recog-	recog-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	39	12	nition	nition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	39	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	39	14	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	39	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	39	16	synthetically	synthetically	RB
work_nvbfnacgujehpi52fk2nc7sg2m	39	17	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	39	18	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	39	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	39	20	degraded	degrade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	39	21	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	39	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	40	1	Finally	finally	RB
work_nvbfnacgujehpi52fk2nc7sg2m	40	2	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	40	3	draw	draw	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	40	4	conclu-	conclu-	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	40	5	sions	sion	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	40	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	40	7	Section	section	NN
work_nvbfnacgujehpi52fk2nc7sg2m	40	8	6	6	CD
work_nvbfnacgujehpi52fk2nc7sg2m	40	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	41	1	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	41	2	Related	related	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	41	3	Work	work	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	4	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	41	5	relation	relation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	6	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	41	7	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	8	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	41	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	41	10	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	11	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	12	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	41	13	attracted	attract	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	41	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	41	15	interest	interest	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	41	17	researchers	researcher	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	41	18	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	41	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	41	20	field	field	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	21	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	41	22	psychology	psychology	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	23	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	41	24	cognitive	cognitive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	41	25	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	41	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	42	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	42	2	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	42	3	similarity	similarity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	42	4	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	42	5	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	42	6	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	42	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	42	8	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	42	9	shown	show	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	42	10	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	42	11	affect	affect	VB
work_nvbfnacgujehpi52fk2nc7sg2m	42	12	human	human	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	42	13	per-	per-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	42	14	formance	formance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	42	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	42	16	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	42	17	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	42	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	42	19	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	43	1	Grainger	Grainger	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	43	2	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	43	3	1990	1990	CD
work_nvbfnacgujehpi52fk2nc7sg2m	43	4	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	43	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	43	6	Andrews	Andrews	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	43	7	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	43	8	1992	1992	CD
work_nvbfnacgujehpi52fk2nc7sg2m	43	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	43	10	1997	1997	CD
work_nvbfnacgujehpi52fk2nc7sg2m	43	11	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	43	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	43	13	Perea	Perea	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	43	14	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	43	15	Rosa	Rosa	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	43	16	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	43	17	2000	2000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	43	18	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	43	19	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	44	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	44	2	commonly	commonly	RB
work_nvbfnacgujehpi52fk2nc7sg2m	44	3	used	use	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	44	4	definition	definition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	44	5	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	44	6	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	44	7	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	44	8	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	44	9	introduced	introduce	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	44	10	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	44	11	Coltheart	Coltheart	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	44	12	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	44	13	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	44	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	45	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	45	2	1977	1977	CD
work_nvbfnacgujehpi52fk2nc7sg2m	45	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	45	4	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	45	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	45	6	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	45	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	45	8	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	45	9	that	that	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	45	10	can	can	MD
work_nvbfnacgujehpi52fk2nc7sg2m	45	11	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	45	12	received	receive	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	45	13	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	45	14	replacing	replace	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	45	15	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	45	16	single	single	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	45	17	letter	letter	NN
work_nvbfnacgujehpi52fk2nc7sg2m	45	18	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	45	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	45	20	original	original	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	45	21	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	45	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	46	1	Recently	recently	RB
work_nvbfnacgujehpi52fk2nc7sg2m	46	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	46	3	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	46	4	more	more	RBR
work_nvbfnacgujehpi52fk2nc7sg2m	46	5	flexible	flexible	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	46	6	definition	definition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	46	7	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	46	8	suggested	suggest	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	46	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	46	10	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	46	11	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	46	12	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	46	13	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	46	14	Levenshtein	Levenshtein	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	46	15	’s	’s	POS
work_nvbfnacgujehpi52fk2nc7sg2m	46	16	string	string	NN
work_nvbfnacgujehpi52fk2nc7sg2m	46	17	distance	distance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	46	18	metric	metric	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	46	19	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	46	20	Levenshtein	Levenshtein	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	46	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	46	22	1966	1966	CD
work_nvbfnacgujehpi52fk2nc7sg2m	46	23	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	46	24	Yarkoni	Yarkoni	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	46	25	et	et	FW
work_nvbfnacgujehpi52fk2nc7sg2m	46	26	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	46	27	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	46	28	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	46	29	2008	2008	CD
work_nvbfnacgujehpi52fk2nc7sg2m	46	30	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	46	31	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	47	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	47	2	new	new	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	47	3	method	method	NN
work_nvbfnacgujehpi52fk2nc7sg2m	47	4	takes	take	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	47	5	into	into	IN
work_nvbfnacgujehpi52fk2nc7sg2m	47	6	account	account	NN
work_nvbfnacgujehpi52fk2nc7sg2m	47	7	letter	letter	NN
work_nvbfnacgujehpi52fk2nc7sg2m	47	8	replace-	replace-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	47	9	ment	ment	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	47	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	47	11	but	but	CC
work_nvbfnacgujehpi52fk2nc7sg2m	47	12	also	also	RB
work_nvbfnacgujehpi52fk2nc7sg2m	47	13	insertion	insertion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	47	14	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	47	15	deletion	deletion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	47	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	48	1	Word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	48	2	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	48	3	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	48	4	also	also	RB
work_nvbfnacgujehpi52fk2nc7sg2m	48	5	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	48	6	shown	show	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	48	7	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	48	8	have	have	VB
work_nvbfnacgujehpi52fk2nc7sg2m	48	9	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	48	10	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	48	11	visual	visual	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	48	12	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	48	13	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	48	14	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	48	15	Richards	Richards	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	48	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	48	17	Heller	Heller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	48	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	48	19	1976	1976	CD
work_nvbfnacgujehpi52fk2nc7sg2m	48	20	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	48	21	New	New	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	48	22	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	48	23	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	48	24	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	48	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	48	26	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	48	27	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	48	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	49	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	49	2	segmentation	segmentation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	49	3	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	49	4	free	free	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	49	5	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	49	6	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	49	7	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	49	8	script	script	NN
work_nvbfnacgujehpi52fk2nc7sg2m	49	9	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	49	10	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	49	11	become	become	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	49	12	more	more	RBR
work_nvbfnacgujehpi52fk2nc7sg2m	49	13	popular	popular	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	49	14	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	49	15	recent	recent	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	49	16	years	year	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	49	17	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	49	18	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	50	1	Madhvanath	Madhvanath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	50	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	50	3	Govindaraju	Govindaraju	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	50	4	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	50	5	2001	2001	CD
work_nvbfnacgujehpi52fk2nc7sg2m	50	6	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	50	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	51	1	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	51	2	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	51	3	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	51	4	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	51	5	adapted	adapt	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	51	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	51	7	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	51	8	recogni-	recogni-	XX
work_nvbfnacgujehpi52fk2nc7sg2m	51	9	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	51	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	51	11	historical	historical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	51	12	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	51	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	51	14	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	51	15	usually	usually	RB
work_nvbfnacgujehpi52fk2nc7sg2m	51	16	suffer	suffer	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	51	17	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	51	18	high	high	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	51	19	level	level	NN
work_nvbfnacgujehpi52fk2nc7sg2m	51	20	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	51	21	degradation	degradation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	51	22	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	51	23	Lavrenko	Lavrenko	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	51	24	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	51	25	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	51	26	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	51	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	51	28	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	51	29	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	51	30	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	52	1	Spitz	Spitz	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	52	2	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	52	3	1999	1999	CD
work_nvbfnacgujehpi52fk2nc7sg2m	52	4	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	52	5	makes	make	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	52	6	use	use	NN
work_nvbfnacgujehpi52fk2nc7sg2m	52	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	52	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	52	9	presence	presence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	52	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	52	11	ascending	ascend	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	52	12	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	52	13	descending	descend	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	52	14	letters	letter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	52	15	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	52	16	both	both	CC
work_nvbfnacgujehpi52fk2nc7sg2m	52	17	recogni-	recogni-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	52	18	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	52	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	52	20	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	52	21	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	52	22	identification	identification	NN
work_nvbfnacgujehpi52fk2nc7sg2m	52	23	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	52	24	Spitz	Spitz	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	52	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	52	26	1997	1997	CD
work_nvbfnacgujehpi52fk2nc7sg2m	52	27	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	52	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	53	1	Some	some	DT
work_nvbfnacgujehpi52fk2nc7sg2m	53	2	work	work	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	3	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	53	4	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	53	5	done	do	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	53	6	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	53	7	image	image	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	8	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	9	meas-	meas-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	10	urement	urement	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	11	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	53	12	OCR	OCR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	53	13	result	result	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	14	prediction	prediction	NN
work_nvbfnacgujehpi52fk2nc7sg2m	53	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	53	16	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	54	1	Esakov	Esakov	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	54	2	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	54	3	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	54	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	55	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	55	2	1994	1994	CD
work_nvbfnacgujehpi52fk2nc7sg2m	55	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	55	4	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	55	5	Blando	Blando	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	55	6	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	55	7	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	55	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	56	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	56	2	1995	1995	CD
work_nvbfnacgujehpi52fk2nc7sg2m	56	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	56	4	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	56	5	Ye	Ye	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	56	6	et	et	FW
work_nvbfnacgujehpi52fk2nc7sg2m	56	7	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	56	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	57	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	57	2	2012	2012	CD
work_nvbfnacgujehpi52fk2nc7sg2m	57	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	57	4	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	57	5	Salah	Salah	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	57	6	et	et	FW
work_nvbfnacgujehpi52fk2nc7sg2m	57	7	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	57	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	58	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	58	2	2013	2013	CD
work_nvbfnacgujehpi52fk2nc7sg2m	58	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	58	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	58	5	but	but	CC
work_nvbfnacgujehpi52fk2nc7sg2m	58	6	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	58	7	mostly	mostly	RB
work_nvbfnacgujehpi52fk2nc7sg2m	58	8	focus	focus	NN
work_nvbfnacgujehpi52fk2nc7sg2m	58	9	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	58	10	OCR	ocr	NN
work_nvbfnacgujehpi52fk2nc7sg2m	58	11	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	58	12	our	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	58	13	work	work	NN
work_nvbfnacgujehpi52fk2nc7sg2m	58	14	deals	deal	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	58	15	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	58	16	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	58	17	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	58	18	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	58	19	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	59	1	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	59	2	Natural	Natural	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	59	3	Language	Language	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	59	4	Characteristics	Characteristics	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	59	5	Below	below	IN
work_nvbfnacgujehpi52fk2nc7sg2m	59	6	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	59	7	present	present	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	59	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	59	9	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	59	10	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	59	11	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	59	12	use	use	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	59	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	59	14	our	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	59	15	hypotheses	hypothesis	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	59	16	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	59	17	their	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	59	18	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	59	19	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	59	20	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	59	21	per-	per-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	59	22	formance	formance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	59	23	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	59	24	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	59	25	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	59	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	60	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	60	2	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	60	3	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	60	4	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	60	5	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	60	6	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	60	7	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	60	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	60	9	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	60	10	neighbor-	neighbor-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	60	11	hood	hood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	60	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	60	13	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	60	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	60	15	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	60	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	60	17	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	60	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	61	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	2	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	61	3	preliminary	preliminary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	61	4	stage	stage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	61	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	61	6	examined	examine	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	61	7	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	61	8	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	61	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	10	possible	possible	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	61	11	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	61	12	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	61	13	decided	decide	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	61	14	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	61	15	focus	focus	VB
work_nvbfnacgujehpi52fk2nc7sg2m	61	16	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	61	18	charac-	charac-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	61	19	teristics	teristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	61	20	above	above	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	21	due	due	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	22	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	23	their	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	61	24	high	high	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	61	25	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	61	26	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	27	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	61	28	structural	structural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	61	29	shape	shape	NN
work_nvbfnacgujehpi52fk2nc7sg2m	61	30	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	61	31	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	61	32	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	62	1	Throughout	throughout	IN
work_nvbfnacgujehpi52fk2nc7sg2m	62	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	62	3	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	62	4	article	article	NN
work_nvbfnacgujehpi52fk2nc7sg2m	62	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	62	6	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	62	7	use	use	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	62	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	62	9	same	same	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	62	10	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	62	11	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	62	12	all	all	PDT
work_nvbfnacgujehpi52fk2nc7sg2m	62	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	62	14	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	62	15	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	62	16	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	62	17	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	62	18	described	describe	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	62	19	below	below	RB
work_nvbfnacgujehpi52fk2nc7sg2m	62	20	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	62	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	63	1	O.	O.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	63	2	Biller	Biller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	63	3	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	63	4	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	63	5	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	64	1	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	64	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	64	3	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	64	4	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	64	6	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	7	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	64	9	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	64	10	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	64	11	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	12	en	en	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	13	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	14	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	15	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	16	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	17	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	64	18	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	19	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	20	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	21	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	22	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	64	23	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	24	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	25	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	26	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	64	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	64	28	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	64	29	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	30	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	31	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	32	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	64	33	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	64	34	paper	paper	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	35	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	64	36	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	64	37	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	64	38	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	64	39	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	64	40	paper	paper	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	41	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	64	42	3.1	3.1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	64	43	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	64	44	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	45	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	64	46	expect	expect	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	64	47	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	48	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	64	49	algorithms	algorithm	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	64	50	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	64	51	perform	perform	VB
work_nvbfnacgujehpi52fk2nc7sg2m	64	52	better	well	RBR
work_nvbfnacgujehpi52fk2nc7sg2m	64	53	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	64	54	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	64	55	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	64	56	longer	long	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	64	57	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	64	58	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	65	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	65	2	consider	consider	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	65	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	65	4	averages	average	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	65	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	65	6	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	65	7	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	65	8	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	65	9	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	65	10	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	65	11	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	65	12	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	65	13	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	65	14	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	65	15	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	66	1	3.2	3.2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	66	2	Orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	66	3	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	66	4	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	66	5	characteristic	characteristic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	66	6	aims	aim	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	66	7	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	66	8	measure	measure	VB
work_nvbfnacgujehpi52fk2nc7sg2m	66	9	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	66	10	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	66	11	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	66	12	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	66	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	66	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	66	15	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	66	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	67	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	67	2	assump-	assump-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	67	3	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	67	4	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	67	5	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	67	6	low	low	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	67	7	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	67	8	distance	distance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	67	9	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	67	10	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	67	11	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	67	12	manifested	manifest	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	67	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	67	14	small	small	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	67	15	distance	distance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	67	16	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	67	17	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	67	18	images	image	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	67	19	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	67	20	contributes	contribute	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	67	21	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	67	22	lower	lower	VB
work_nvbfnacgujehpi52fk2nc7sg2m	67	23	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	67	24	rates	rate	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	67	25	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	68	1	An	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	68	2	accepted	accepted	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	68	3	measure	measure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	68	4	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	68	5	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	68	6	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	68	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	68	8	ortho-	ortho-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	68	9	graphic	graphic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	68	10	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	68	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	69	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	69	2	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	69	3	article	article	NN
work_nvbfnacgujehpi52fk2nc7sg2m	69	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	69	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	69	6	use	use	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	69	7	Levenshtein	Levenshtein	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	69	8	distance	distance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	69	9	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	69	10	determining	determine	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	69	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	69	12	ortho-	ortho-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	69	13	graphic	graphic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	69	14	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	69	15	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	70	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	70	2	take	take	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	70	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	70	4	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	70	5	neigh-	neigh-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	70	6	borhood	borhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	70	7	size	size	NN
work_nvbfnacgujehpi52fk2nc7sg2m	70	8	over	over	IN
work_nvbfnacgujehpi52fk2nc7sg2m	70	9	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	70	10	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	70	11	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	70	12	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	70	13	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	70	14	representative	representative	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	70	15	measure	measure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	70	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	71	1	3.3	3.3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	71	2	Ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	71	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	71	4	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	71	5	An	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	71	6	important	important	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	71	7	aspect	aspect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	8	that	that	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	71	9	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	71	10	high	high	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	71	11	impact	impact	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	12	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	71	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	71	14	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	71	16	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	17	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	18	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	71	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	71	20	graphical	graphical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	71	21	proper-	proper-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	71	22	ties	tie	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	71	23	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	71	24	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	71	25	character	character	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	26	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	71	27	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	72	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	72	2	focus	focus	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	72	3	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	72	4	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	72	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	72	6	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	72	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	73	1	Graphically	graphically	RB
work_nvbfnacgujehpi52fk2nc7sg2m	73	2	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	73	3	handwritten	handwritten	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	73	4	line	line	NN
work_nvbfnacgujehpi52fk2nc7sg2m	73	5	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	73	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	73	7	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	73	8	baseline	baseline	NN
work_nvbfnacgujehpi52fk2nc7sg2m	73	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	73	10	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	73	11	upper	upper	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	73	12	baseline	baseline	NN
work_nvbfnacgujehpi52fk2nc7sg2m	73	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	73	14	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	73	15	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	73	16	most	most	JJS
work_nvbfnacgujehpi52fk2nc7sg2m	73	17	letters	letter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	73	18	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	73	19	written	write	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	73	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	74	1	Some	some	DT
work_nvbfnacgujehpi52fk2nc7sg2m	74	2	letters	letter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	74	3	exceed	exceed	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	74	4	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	74	5	baselines	baseline	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	74	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	75	1	Those	those	DT
work_nvbfnacgujehpi52fk2nc7sg2m	75	2	that	that	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	75	3	exceed	exceed	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	75	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	75	5	upper	upper	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	75	6	baseline	baseline	NN
work_nvbfnacgujehpi52fk2nc7sg2m	75	7	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	75	8	called	call	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	75	9	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	75	10	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	75	11	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	76	1	b	b	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	76	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	76	3	d	d	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	76	4	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	76	5	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	77	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	77	2	ones	one	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	77	3	that	that	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	77	4	exceed	exceed	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	77	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	77	6	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	77	7	baseline	baseline	NN
work_nvbfnacgujehpi52fk2nc7sg2m	77	8	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	77	9	called	call	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	77	10	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	77	11	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	77	12	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	78	1	g	g	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	78	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	78	3	q	q	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	78	4	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	78	5	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	79	1	Ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	79	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	79	3	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	79	4	influence	influence	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	79	5	dramatically	dramatically	RB
work_nvbfnacgujehpi52fk2nc7sg2m	79	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	79	7	outer	outer	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	79	8	shape	shape	NN
work_nvbfnacgujehpi52fk2nc7sg2m	79	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	79	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	79	11	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	79	12	’s	’s	POS
work_nvbfnacgujehpi52fk2nc7sg2m	79	13	image	image	NN
work_nvbfnacgujehpi52fk2nc7sg2m	79	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	79	15	therefore	therefore	RB
work_nvbfnacgujehpi52fk2nc7sg2m	79	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	79	17	might	may	MD
work_nvbfnacgujehpi52fk2nc7sg2m	79	18	effect	effect	VB
work_nvbfnacgujehpi52fk2nc7sg2m	79	19	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	79	20	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	79	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	80	1	Several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	80	2	approaches	approach	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	80	3	use	use	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	80	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	80	5	presence	presence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	6	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	80	7	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	80	8	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	80	9	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	80	10	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	80	11	recogni-	recogni-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	12	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	80	14	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	15	identification	identification	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	80	17	but	but	CC
work_nvbfnacgujehpi52fk2nc7sg2m	80	18	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	80	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	80	20	best	good	JJS
work_nvbfnacgujehpi52fk2nc7sg2m	80	21	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	80	22	our	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	80	23	knowledge	knowledge	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	80	25	no	no	DT
work_nvbfnacgujehpi52fk2nc7sg2m	80	26	work	work	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	27	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	80	28	been	be	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	80	29	done	do	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	80	30	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	80	31	quantify	quantify	VB
work_nvbfnacgujehpi52fk2nc7sg2m	80	32	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	80	33	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	80	34	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	81	1	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	81	2	Distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	81	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	81	4	Orthographic	Orthographic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	81	5	Characteristics	Characteristics	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	81	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	81	7	Natural	Natural	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	81	8	Languages	Languages	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	81	9	Here	here	RB
work_nvbfnacgujehpi52fk2nc7sg2m	81	10	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	81	11	analyze	analyze	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	81	12	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	81	13	behavior	behavior	NN
work_nvbfnacgujehpi52fk2nc7sg2m	81	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	81	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	81	16	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	81	17	described	describe	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	81	18	above	above	RB
work_nvbfnacgujehpi52fk2nc7sg2m	81	19	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	81	20	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	81	21	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	81	22	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	81	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	82	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	82	2	collected	collect	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	82	3	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	82	4	corpus	corpus	NN
work_nvbfnacgujehpi52fk2nc7sg2m	82	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	82	6	books	book	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	82	7	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	82	8	four	four	CD
work_nvbfnacgujehpi52fk2nc7sg2m	82	9	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	82	10	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	82	11	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	82	12	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	82	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	82	14	Russian	russian	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	82	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	82	16	Hebrew	Hebrew	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	82	17	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	82	18	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	82	19	Arabic	arabic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	82	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	83	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	83	2	books	book	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	83	3	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	83	4	used	use	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	83	5	were	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	83	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	83	7	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	83	8	collection	collection	NN
work_nvbfnacgujehpi52fk2nc7sg2m	83	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	83	10	prose	prose	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	83	11	books	book	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	83	12	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	83	13	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	83	14	format	format	NN
work_nvbfnacgujehpi52fk2nc7sg2m	83	15	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	83	16	some	some	DT
work_nvbfnacgujehpi52fk2nc7sg2m	83	17	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	83	18	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	83	19	were	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	83	20	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	83	21	Adventures	adventure	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	83	22	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	83	23	Huckleberry	Huckleberry	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	83	24	Finn	Finn	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	83	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	83	26	Crime	Crime	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	83	27	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	83	28	Punishment	Punishment	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	83	29	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	83	30	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	83	31	variety	variety	NN
work_nvbfnacgujehpi52fk2nc7sg2m	83	32	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	83	33	short	short	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	83	34	stories	story	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	83	35	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	83	36	were	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	83	37	available	available	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	83	38	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	83	39	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	83	40	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	83	41	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	83	42	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	84	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	84	2	processed	process	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	84	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	84	4	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	84	5	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	84	6	extracting	extract	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	84	7	distinct	distinct	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	84	8	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	84	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	84	10	building	build	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	84	11	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	84	12	dictionary	dictionary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	84	13	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	84	14	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	84	15	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	84	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	85	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	85	2	prefer	prefer	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	85	3	collecting	collect	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	85	4	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	85	5	statistics	statistic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	85	6	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	85	7	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	85	8	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	85	9	not	not	RB
work_nvbfnacgujehpi52fk2nc7sg2m	85	10	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	85	11	dic-	dic-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	85	12	tionaries	tionarie	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	85	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	85	14	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	85	15	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	85	16	get	get	VB
work_nvbfnacgujehpi52fk2nc7sg2m	85	17	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	85	18	better	well	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	85	19	representation	representation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	85	20	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	85	21	frequently	frequently	RB
work_nvbfnacgujehpi52fk2nc7sg2m	85	22	used	use	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	85	23	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	85	24	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	85	25	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	85	26	ignore	ignore	VB
work_nvbfnacgujehpi52fk2nc7sg2m	85	27	rare	rare	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	85	28	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	85	29	unused	unused	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	85	30	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	85	31	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	86	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	86	2	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	86	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	86	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	86	5	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	86	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	86	7	our	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	86	8	database	database	NN
work_nvbfnacgujehpi52fk2nc7sg2m	86	9	contains	contain	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	86	10	about	about	IN
work_nvbfnacgujehpi52fk2nc7sg2m	86	11	18,000	18,000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	86	12	unique	unique	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	86	13	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	86	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	87	1	4.1	4.1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	87	2	Word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	87	3	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	87	4	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	87	5	Figure	figure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	87	6	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	87	7	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	87	8	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	87	9	both	both	CC
work_nvbfnacgujehpi52fk2nc7sg2m	87	10	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	87	11	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	87	12	Russian	Russian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	87	13	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	87	14	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	87	15	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	87	16	longer	long	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	87	17	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	87	18	than	than	IN
work_nvbfnacgujehpi52fk2nc7sg2m	87	19	Hebrew	hebrew	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	87	20	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	87	21	Arabic	Arabic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	87	22	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	87	23	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	87	24	wider	wide	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	87	25	spread	spread	NN
work_nvbfnacgujehpi52fk2nc7sg2m	87	26	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	87	27	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	87	28	lengths	length	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	87	29	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	88	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	88	2	graph	graph	NN
work_nvbfnacgujehpi52fk2nc7sg2m	88	3	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	4	Figure	Figure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	88	5	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	88	6	illustrates	illustrate	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	88	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	88	8	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	88	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	10	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	88	11	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	88	12	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	13	showing	show	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	88	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	88	15	percentage	percentage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	88	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	17	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	88	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	88	19	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	20	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	88	21	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	88	22	over	over	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	23	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	88	24	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	88	25	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	88	26	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	88	27	corpus	corpus	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	88	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	89	1	It	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	89	2	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	89	3	substantial	substantial	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	89	4	differences	difference	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	89	5	among	among	IN
work_nvbfnacgujehpi52fk2nc7sg2m	89	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	89	7	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	89	8	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	89	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	90	1	4.2	4.2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	90	2	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	90	3	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	90	4	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	90	5	To	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	90	6	evaluate	evaluate	VB
work_nvbfnacgujehpi52fk2nc7sg2m	90	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	8	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	90	9	density	density	NN
work_nvbfnacgujehpi52fk2nc7sg2m	90	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	90	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	12	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	90	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	90	14	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	15	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	90	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	90	17	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	90	18	calculate	calculate	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	90	19	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	90	20	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	21	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	90	22	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	90	23	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	24	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	90	25	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	26	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	90	27	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	90	28	its	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	90	29	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	90	30	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	90	31	within	within	IN
work_nvbfnacgujehpi52fk2nc7sg2m	90	32	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	90	33	corpus	corpus	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	90	34	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	90	35	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	91	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	91	2	graph	graph	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	3	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	91	4	Figure	Figure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	91	5	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	91	6	shows	show	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	91	7	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	91	8	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	91	9	lan-	lan-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	10	guage	guage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	91	12	percentage	percentage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	13	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	91	14	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	91	15	having	have	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	91	16	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	91	17	specific	specific	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	91	18	neighbor	neighbor	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	19	count	count	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	20	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	91	21	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	91	22	x	x	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	23	-	-	:
work_nvbfnacgujehpi52fk2nc7sg2m	91	24	axis	axis	NN
work_nvbfnacgujehpi52fk2nc7sg2m	91	25	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	91	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	92	1	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	2	can	can	MD
work_nvbfnacgujehpi52fk2nc7sg2m	92	3	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	92	4	seen	see	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	92	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	92	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	7	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	92	8	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	92	9	Russian	Russian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	92	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	92	11	main	main	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	92	12	mass	mass	NN
work_nvbfnacgujehpi52fk2nc7sg2m	92	13	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	14	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	92	15	con-	con-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	92	16	centrates	centrate	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	92	17	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	18	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	92	19	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	92	20	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	92	21	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	22	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	92	23	per	per	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	24	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	92	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	92	26	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	27	Hebrew	hebrew	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	92	28	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	92	29	Arabic	Arabic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	92	30	spread	spread	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	92	31	also	also	RB
work_nvbfnacgujehpi52fk2nc7sg2m	92	32	over	over	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	33	higher	high	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	92	34	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	92	35	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	92	36	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	92	37	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	93	1	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	93	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	94	1	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	2	Distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	4	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	5	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	6	per	per	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	7	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	8	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	94	9	percentage	percentage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	11	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	94	12	per	per	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	13	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	14	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	15	Language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	16	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	94	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	94	18	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	19	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	20	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	94	21	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	94	22	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	94	23	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	94	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	94	25	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	26	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	27	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	28	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	29	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	30	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	31	en	en	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	32	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	94	33	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	34	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	94	35	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	36	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	94	37	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	38	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	39	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	40	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	41	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	42	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	94	43	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	44	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	45	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	46	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	94	47	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	48	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	49	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	50	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	51	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	94	52	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	53	paper	paper	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	54	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	94	55	'	'	''
work_nvbfnacgujehpi52fk2nc7sg2m	94	56	utilize	utilize	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	57	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	94	58	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	VB
work_nvbfnacgujehpi52fk2nc7sg2m	94	59	4.3	4.3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	94	60	Ascending	ascending	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	61	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	94	62	descending	descend	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	94	63	letters	letter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	94	64	To	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	94	65	measure	measure	VB
work_nvbfnacgujehpi52fk2nc7sg2m	94	66	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	94	67	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	68	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	94	69	ascending	ascending	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	70	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	94	71	des-	des-	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	94	72	cending	cende	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	94	73	letters	letter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	94	74	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	94	75	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	94	76	define	define	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	94	77	ascender	ascender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	78	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	94	79	descender	descender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	80	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	94	81	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	95	1	Each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	95	2	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	95	3	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	95	4	represented	represent	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	95	5	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	95	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	95	7	sequence	sequence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	95	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	95	9	letter	letter	NN
work_nvbfnacgujehpi52fk2nc7sg2m	95	10	types	type	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	95	11	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	95	12	ascending	ascend	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	95	13	/	/	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	95	14	descending/	descending/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	95	15	none	none	NN
work_nvbfnacgujehpi52fk2nc7sg2m	95	16	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	95	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	96	1	Each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	96	2	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	96	3	matches	match	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	96	4	some	some	DT
work_nvbfnacgujehpi52fk2nc7sg2m	96	5	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	96	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	97	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	97	2	example	example	NN
work_nvbfnacgujehpi52fk2nc7sg2m	97	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	97	4	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	97	5	‘	'	``
work_nvbfnacgujehpi52fk2nc7sg2m	97	6	dog	dog	NN
work_nvbfnacgujehpi52fk2nc7sg2m	97	7	’	'	''
work_nvbfnacgujehpi52fk2nc7sg2m	97	8	will	will	MD
work_nvbfnacgujehpi52fk2nc7sg2m	97	9	match	match	VB
work_nvbfnacgujehpi52fk2nc7sg2m	97	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	97	11	config-	config-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	97	12	uration	uration	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	97	13	ascend	ascend	NN
work_nvbfnacgujehpi52fk2nc7sg2m	97	14	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	97	15	none	none	NN
work_nvbfnacgujehpi52fk2nc7sg2m	97	16	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	97	17	descend	descend	NN
work_nvbfnacgujehpi52fk2nc7sg2m	97	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	98	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	98	2	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	98	3	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	98	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	98	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	98	6	compute	compute	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	98	7	how	how	WRB
work_nvbfnacgujehpi52fk2nc7sg2m	98	8	many	many	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	98	9	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	98	10	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	98	11	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	98	12	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	98	13	given	give	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	98	14	con-	con-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	98	15	figuration	figuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	98	16	set	set	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	98	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	99	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	99	2	Table	table	NN
work_nvbfnacgujehpi52fk2nc7sg2m	99	3	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	99	4	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	99	5	present	present	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	99	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	99	7	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	99	8	size	size	NN
work_nvbfnacgujehpi52fk2nc7sg2m	99	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	99	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	99	11	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	99	12	sets	set	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	99	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	99	14	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	99	15	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	99	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	100	1	A	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	100	2	con-	con-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	100	3	figuration	figuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	100	4	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	100	5	considered	consider	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	100	6	empty	empty	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	100	7	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	100	8	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	100	9	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	100	10	if	if	IN
work_nvbfnacgujehpi52fk2nc7sg2m	100	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	100	12	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	100	13	does	do	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	100	14	not	not	RB
work_nvbfnacgujehpi52fk2nc7sg2m	100	15	contain	contain	VB
work_nvbfnacgujehpi52fk2nc7sg2m	100	16	any	any	DT
work_nvbfnacgujehpi52fk2nc7sg2m	100	17	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	100	18	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	100	19	that	that	DT
work_nvbfnacgujehpi52fk2nc7sg2m	100	20	con-	con-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	100	21	figuration	figuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	100	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	101	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	101	2	expect	expect	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	101	3	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	101	4	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	101	5	many	many	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	6	small	small	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	7	non	non	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	8	-	-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	9	empty	empty	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	10	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	101	11	sets	set	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	101	12	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	101	13	have	have	VB
work_nvbfnacgujehpi52fk2nc7sg2m	101	14	more	more	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	101	15	graph-	graph-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	16	ical	ical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	17	diversity	diversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	101	18	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	101	19	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	101	20	’	’	POS
work_nvbfnacgujehpi52fk2nc7sg2m	101	21	general	general	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	101	22	shapes	shape	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	101	23	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	101	24	therefore	therefore	RB
work_nvbfnacgujehpi52fk2nc7sg2m	101	25	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	101	26	have	have	VB
work_nvbfnacgujehpi52fk2nc7sg2m	101	27	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	101	28	advantage	advantage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	101	29	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	101	30	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	101	31	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	101	32	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	102	1	4.4	4.4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	102	2	Summary	Summary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	102	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	102	5	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	102	8	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	9	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	102	11	four	four	CD
work_nvbfnacgujehpi52fk2nc7sg2m	102	12	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	102	13	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	14	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	15	seen	see	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	102	16	above	above	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	17	there	there	EX
work_nvbfnacgujehpi52fk2nc7sg2m	102	18	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	102	19	substantial	substantial	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	102	20	differences	difference	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	21	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	22	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	102	23	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	24	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	25	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	102	26	checked	check	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	102	27	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	28	amongst	amongst	IN
work_nvbfnacgujehpi52fk2nc7sg2m	102	29	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	102	30	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	102	31	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	103	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	103	2	general	general	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	103	3	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	103	4	English	english	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	103	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	103	6	Russian	Russian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	103	7	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	103	8	longer	long	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	103	9	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	103	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	103	11	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	103	12	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	103	13	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	103	14	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	103	15	per	per	IN
work_nvbfnacgujehpi52fk2nc7sg2m	103	16	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	103	17	than	than	IN
work_nvbfnacgujehpi52fk2nc7sg2m	103	18	Hebrew	hebrew	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	103	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	103	20	Arabic	Arabic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	103	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	104	1	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	104	2	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	104	3	ascender	ascender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	104	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	104	5	descender	descender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	104	6	distributions	distribution	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	104	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	104	8	Hebrew	hebrew	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	104	9	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	104	10	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	104	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	104	12	lowest	low	JJS
work_nvbfnacgujehpi52fk2nc7sg2m	104	13	diversity	diversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	104	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	104	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	104	16	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	104	17	has	have	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	104	18	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	104	19	widest	wide	JJS
work_nvbfnacgujehpi52fk2nc7sg2m	104	20	spread	spread	NN
work_nvbfnacgujehpi52fk2nc7sg2m	104	21	among	among	IN
work_nvbfnacgujehpi52fk2nc7sg2m	104	22	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	104	23	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	104	24	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	104	25	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	105	1	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	105	2	Experimental	experimental	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	105	3	Evaluation	Evaluation	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	105	4	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	105	5	previous	previous	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	105	6	sections	section	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	105	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	105	8	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	105	9	listed	list	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	105	10	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	105	11	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	105	12	lan-	lan-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	105	13	guage	guage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	105	14	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	105	15	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	105	16	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	105	17	conjecture	conjecture	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	105	18	influence	influence	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	105	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	105	20	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	105	21	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	105	22	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	105	23	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	105	24	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	106	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	106	2	measured	measure	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	106	3	some	some	DT
work_nvbfnacgujehpi52fk2nc7sg2m	106	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	106	5	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	106	6	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	106	7	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	106	8	corpora	corpora	NN
work_nvbfnacgujehpi52fk2nc7sg2m	106	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	106	10	four	four	CD
work_nvbfnacgujehpi52fk2nc7sg2m	106	11	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	106	12	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	106	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	106	14	found	find	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	106	15	out	out	RP
work_nvbfnacgujehpi52fk2nc7sg2m	106	16	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	106	17	there	there	EX
work_nvbfnacgujehpi52fk2nc7sg2m	106	18	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	106	19	substantial	substantial	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	106	20	differences	difference	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	106	21	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	106	22	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	106	23	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	106	24	among	among	IN
work_nvbfnacgujehpi52fk2nc7sg2m	106	25	dif-	dif-	DT
work_nvbfnacgujehpi52fk2nc7sg2m	106	26	ferent	ferent	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	106	27	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	106	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	107	1	Yet	yet	CC
work_nvbfnacgujehpi52fk2nc7sg2m	107	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	107	3	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	107	4	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	107	5	not	not	RB
work_nvbfnacgujehpi52fk2nc7sg2m	107	6	given	give	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	107	7	any	any	DT
work_nvbfnacgujehpi52fk2nc7sg2m	107	8	estima-	estima-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	107	9	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	107	10	or	or	CC
work_nvbfnacgujehpi52fk2nc7sg2m	107	11	measurement	measurement	NN
work_nvbfnacgujehpi52fk2nc7sg2m	107	12	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	107	13	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	107	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	107	15	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	107	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	107	17	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	107	18	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	107	19	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	107	20	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	107	21	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	107	22	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	107	23	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	107	24	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	108	1	Using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	108	2	real	real	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	108	3	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	108	4	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	108	5	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	108	6	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	108	7	does	do	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	108	8	not	not	RB
work_nvbfnacgujehpi52fk2nc7sg2m	108	9	enable	enable	VB
work_nvbfnacgujehpi52fk2nc7sg2m	108	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	108	11	isolation	isolation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	108	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	108	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	108	14	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	108	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	108	16	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	108	17	characteristic	characteristic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	108	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	109	1	To	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	109	2	measure	measure	VB
work_nvbfnacgujehpi52fk2nc7sg2m	109	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	109	4	impact	impact	NN
work_nvbfnacgujehpi52fk2nc7sg2m	109	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	109	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	109	7	certain	certain	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	109	8	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	109	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	109	10	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	109	11	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	109	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	109	13	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	109	14	generated	generate	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	109	15	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	109	16	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	109	17	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	109	18	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	109	19	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	109	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	109	21	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	109	22	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	109	23	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	109	24	English	english	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	109	25	alphabet	alphabet	NN
work_nvbfnacgujehpi52fk2nc7sg2m	109	26	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	109	27	simi-	simi-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	109	28	lar	lar	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	109	29	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	109	30	every	every	DT
work_nvbfnacgujehpi52fk2nc7sg2m	109	31	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	109	32	but	but	CC
work_nvbfnacgujehpi52fk2nc7sg2m	109	33	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	109	34	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	109	35	one	one	CD
work_nvbfnacgujehpi52fk2nc7sg2m	109	36	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	110	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	110	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	110	3	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	110	4	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	110	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	110	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	110	7	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	110	8	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	110	9	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	110	10	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	110	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	111	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	111	2	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	111	3	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	111	4	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	111	5	generate	generate	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	111	6	docu-	docu-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	111	7	ments	ment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	111	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	111	9	perform	perform	VB
work_nvbfnacgujehpi52fk2nc7sg2m	111	10	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	111	11	spotting	spotting	NN
work_nvbfnacgujehpi52fk2nc7sg2m	111	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	111	13	evaluate	evaluate	VB
work_nvbfnacgujehpi52fk2nc7sg2m	111	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	111	15	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	111	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	111	17	compare	compare	VB
work_nvbfnacgujehpi52fk2nc7sg2m	111	18	their	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	111	19	accuracy	accuracy	NN
work_nvbfnacgujehpi52fk2nc7sg2m	111	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	112	1	Because	because	IN
work_nvbfnacgujehpi52fk2nc7sg2m	112	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	112	3	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	112	4	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	112	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	112	6	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	112	7	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	112	8	similar	similar	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	112	9	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	112	10	every	every	DT
work_nvbfnacgujehpi52fk2nc7sg2m	112	11	aspect	aspect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	112	12	except	except	IN
work_nvbfnacgujehpi52fk2nc7sg2m	112	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	112	14	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	112	15	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	112	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	112	17	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	112	18	distinctive	distinctive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	112	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	112	20	stable	stable	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	112	21	variance	variance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	112	22	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	112	23	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	112	24	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	112	25	would	would	MD
work_nvbfnacgujehpi52fk2nc7sg2m	112	26	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	112	27	confidently	confidently	RB
work_nvbfnacgujehpi52fk2nc7sg2m	112	28	assigned	assign	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	112	29	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	112	30	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	112	31	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	112	32	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	112	33	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	113	1	On	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	113	2	clean	clean	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	113	3	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	113	4	high	high	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	113	5	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	113	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	113	7	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	113	8	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	113	9	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	113	10	would	would	MD
work_nvbfnacgujehpi52fk2nc7sg2m	113	11	receive	receive	VB
work_nvbfnacgujehpi52fk2nc7sg2m	113	12	near	near	IN
work_nvbfnacgujehpi52fk2nc7sg2m	113	13	perfect	perfect	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	113	14	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	113	15	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	113	16	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	113	17	regardless	regardless	RB
work_nvbfnacgujehpi52fk2nc7sg2m	113	18	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	113	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	113	20	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	113	21	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	113	22	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	113	23	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	113	24	characteris-	characteris-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	113	25	tics	tic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	113	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	114	1	Therefore	therefore	RB
work_nvbfnacgujehpi52fk2nc7sg2m	114	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	114	3	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	114	4	added	add	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	114	5	noise	noise	NN
work_nvbfnacgujehpi52fk2nc7sg2m	114	6	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	114	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	114	8	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	114	9	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	114	10	create	create	VB
work_nvbfnacgujehpi52fk2nc7sg2m	114	11	variance	variance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	114	12	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	114	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	114	14	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	114	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	114	16	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	114	17	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	114	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	115	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	115	2	used	use	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	115	3	few	few	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	115	4	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	115	5	combinations	combination	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	115	6	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	115	7	degradation	degradation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	115	8	types	type	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	115	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	115	10	levels	level	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	115	11	such	such	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	115	12	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	115	13	simple	simple	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	115	14	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	115	15	Gaussian	Gaussian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	115	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	115	17	salt	salt	NN
work_nvbfnacgujehpi52fk2nc7sg2m	115	18	&	&	CC
work_nvbfnacgujehpi52fk2nc7sg2m	115	19	pepper	pepper	NN
work_nvbfnacgujehpi52fk2nc7sg2m	115	20	noise	noise	NN
work_nvbfnacgujehpi52fk2nc7sg2m	115	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	115	22	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	115	23	Kanungo	Kanungo	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	115	24	’s	’s	POS
work_nvbfnacgujehpi52fk2nc7sg2m	115	25	degradation	degradation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	115	26	model	model	NN
work_nvbfnacgujehpi52fk2nc7sg2m	115	27	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	115	28	Kanungo	Kanungo	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	115	29	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	115	30	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	115	31	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	115	32	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	115	33	1993	1993	CD
work_nvbfnacgujehpi52fk2nc7sg2m	115	34	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	115	35	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	116	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	116	2	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	116	3	experiments	experiment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	116	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	116	5	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	116	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	117	1	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	2	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	117	3	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	5	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	6	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	7	count	count	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	8	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	117	9	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	117	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	117	11	percentage	percentage	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	13	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	14	having	have	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	117	15	specified	specify	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	117	16	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	17	count	count	VB
work_nvbfnacgujehpi52fk2nc7sg2m	117	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	117	19	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	20	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	117	22	Hebrew	Hebrew	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	23	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	117	24	Russian	Russian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	25	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	117	26	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	117	27	Arabic	Arabic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	28	Table	Table	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	29	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	30	Ascender	ascender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	31	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	117	32	descender	descender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	33	configurations	configuration	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	34	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	117	35	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	36	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	117	37	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	38	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	117	39	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	117	40	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	41	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	42	non	non	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	43	-	-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	44	empty	empty	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	45	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	46	sets	set	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	47	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	117	48	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	117	49	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	50	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	51	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	52	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	53	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	117	54	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	117	55	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	56	set	set	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	117	57	Language	Language	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	58	Non	non	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	59	-	-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	60	empty	empty	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	61	configurations	configuration	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	117	62	Average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	63	configuration	configuration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	64	size	size	NN
work_nvbfnacgujehpi52fk2nc7sg2m	117	65	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	66	4,218	4,218	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	67	4.29	4.29	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	68	Russian	russian	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	69	2,742	2,742	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	70	6.87	6.87	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	71	Hebrew	Hebrew	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	72	499	499	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	73	39.59	39.59	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	74	Arabic	arabic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	117	75	1,560	1,560	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	76	11.77	11.77	CD
work_nvbfnacgujehpi52fk2nc7sg2m	117	77	O.	O.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	78	Biller	Biller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	79	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	80	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	117	81	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	118	1	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	118	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	3	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	118	4	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	118	6	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	7	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	118	9	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	118	10	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	11	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	12	en	en	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	13	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	14	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	15	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	16	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	17	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	118	18	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	19	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	20	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	21	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	22	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	23	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	118	24	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	25	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	26	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	118	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	118	28	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	118	29	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	30	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	31	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	32	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	118	33	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	34	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	35	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	36	t	t	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	37	'	'	''
work_nvbfnacgujehpi52fk2nc7sg2m	118	38	'	'	''
work_nvbfnacgujehpi52fk2nc7sg2m	118	39	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	40	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	41	t	t	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	42	Since	since	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	43	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	44	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	45	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	118	46	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	118	47	reaction	reaction	NN
work_nvbfnacgujehpi52fk2nc7sg2m	118	48	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	118	49	change	change	VB
work_nvbfnacgujehpi52fk2nc7sg2m	118	50	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	51	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	118	52	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	118	53	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	118	54	similar	similar	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	118	55	across	across	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	56	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	118	57	types	type	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	118	58	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	118	59	degradations	degradation	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	118	60	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	119	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	119	2	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	119	3	provided	provide	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	119	4	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	119	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	119	6	averages	average	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	119	7	over	over	IN
work_nvbfnacgujehpi52fk2nc7sg2m	119	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	119	9	used	use	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	119	10	deg-	deg-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	119	11	radation	radation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	119	12	models	model	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	119	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	120	1	Example	example	NN
work_nvbfnacgujehpi52fk2nc7sg2m	120	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	120	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	120	4	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	120	5	noise	noise	NN
work_nvbfnacgujehpi52fk2nc7sg2m	120	6	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	120	7	shown	show	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	120	8	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	120	9	Figure	Figure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	120	10	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	120	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	121	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	121	2	degradation	degradation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	121	3	models	model	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	121	4	’	’	POS
work_nvbfnacgujehpi52fk2nc7sg2m	121	5	param-	param-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	121	6	etrization	etrization	NN
work_nvbfnacgujehpi52fk2nc7sg2m	121	7	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	121	8	adjusted	adjust	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	121	9	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	121	10	receive	receive	VB
work_nvbfnacgujehpi52fk2nc7sg2m	121	11	mediocre	mediocre	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	121	12	recogni-	recogni-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	121	13	tion	tion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	121	14	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	121	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	121	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	121	17	stayed	stay	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	121	18	constant	constant	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	121	19	throughout	throughout	IN
work_nvbfnacgujehpi52fk2nc7sg2m	121	20	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	121	21	experiments	experiment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	121	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	122	1	Among	among	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	122	3	many	many	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	122	4	available	available	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	122	5	approaches	approach	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	122	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	7	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	122	8	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	122	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	122	10	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	122	11	chose	choose	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	122	12	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	122	13	start	start	VB
work_nvbfnacgujehpi52fk2nc7sg2m	122	14	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	15	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	16	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	122	17	approach	approach	NN
work_nvbfnacgujehpi52fk2nc7sg2m	122	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	122	19	presented	present	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	122	20	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	21	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	22	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	122	23	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	24	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	122	25	2003b	2003b	CD
work_nvbfnacgujehpi52fk2nc7sg2m	122	26	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	122	27	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	28	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	122	29	combination	combination	NN
work_nvbfnacgujehpi52fk2nc7sg2m	122	30	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	31	profile	profile	NN
work_nvbfnacgujehpi52fk2nc7sg2m	122	32	features	feature	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	122	33	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	122	34	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	35	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	122	36	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	37	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	122	38	2003a	2003a	CD
work_nvbfnacgujehpi52fk2nc7sg2m	122	39	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	122	40	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	122	41	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	122	42	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	122	43	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	44	method	method	NN
work_nvbfnacgujehpi52fk2nc7sg2m	122	45	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	122	46	Zhang	Zhang	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	47	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	48	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	49	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	122	50	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	122	51	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	122	52	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	122	53	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	123	1	5.1	5.1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	123	2	Evaluation	evaluation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	123	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	123	5	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	6	spotting	spot	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	123	7	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	123	8	part	part	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	123	10	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	11	generation	generation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	123	13	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	123	14	construct	construct	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	123	15	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	123	16	ground	ground	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	17	truth	truth	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	18	data	datum	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	123	19	base	base	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	20	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	123	21	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	123	22	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	123	23	test	test	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	123	24	our	-PRON-	PRP$
work_nvbfnacgujehpi52fk2nc7sg2m	123	25	rec-	rec-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	26	ognition	ognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	123	27	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	123	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	124	1	Using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	124	2	both	both	DT
work_nvbfnacgujehpi52fk2nc7sg2m	124	3	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	124	4	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	124	5	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	124	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	124	7	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	124	8	retrieve	retrieve	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	124	9	all	all	PDT
work_nvbfnacgujehpi52fk2nc7sg2m	124	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	124	11	occurrences	occurrence	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	124	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	124	13	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	124	14	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	124	15	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	124	16	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	124	17	corpus	corpus	NN
work_nvbfnacgujehpi52fk2nc7sg2m	124	18	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	124	19	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	124	20	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	124	21	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	124	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	125	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	125	2	overall	overall	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	125	3	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	125	4	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	125	5	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	125	6	given	give	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	125	7	method	method	NN
work_nvbfnacgujehpi52fk2nc7sg2m	125	8	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	125	9	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	125	10	given	give	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	125	11	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	125	12	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	125	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	125	14	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	125	15	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	125	16	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	125	17	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	125	18	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	125	19	all	all	PDT
work_nvbfnacgujehpi52fk2nc7sg2m	125	20	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	125	21	retrieved	retrieve	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	125	22	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	125	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	126	1	5.2	5.2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	126	2	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	126	3	testing	testing	NN
work_nvbfnacgujehpi52fk2nc7sg2m	126	4	system	system	NN
work_nvbfnacgujehpi52fk2nc7sg2m	126	5	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	126	6	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	126	7	built	build	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	126	8	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	126	9	testing	testing	NN
work_nvbfnacgujehpi52fk2nc7sg2m	126	10	system	system	NN
work_nvbfnacgujehpi52fk2nc7sg2m	126	11	that	that	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	126	12	runs	run	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	126	13	sets	set	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	126	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	126	15	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	126	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	127	1	Each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	127	2	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	127	3	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	127	4	composed	compose	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	127	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	127	6	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	127	7	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	127	8	similar	similar	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	127	9	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	127	10	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	127	11	differ	differ	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	127	12	only	only	RB
work_nvbfnacgujehpi52fk2nc7sg2m	127	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	127	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	127	15	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	127	16	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	127	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	128	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	128	2	process	process	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	3	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	128	4	one	one	CD
work_nvbfnacgujehpi52fk2nc7sg2m	128	5	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	6	includes	include	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	128	7	creation	creation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	128	9	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	128	10	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	11	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	128	12	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	128	13	noised	noised	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	128	14	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	15	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	128	16	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	128	17	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	128	19	running	run	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	128	20	two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	128	21	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	22	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	23	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	128	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	128	25	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	128	26	evalu-	evalu-	XX
work_nvbfnacgujehpi52fk2nc7sg2m	128	27	ating	ate	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	128	28	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	128	29	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	128	30	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	128	31	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	128	32	both	both	DT
work_nvbfnacgujehpi52fk2nc7sg2m	128	33	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	129	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	129	2	whole	whole	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	129	3	process	process	NN
work_nvbfnacgujehpi52fk2nc7sg2m	129	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	129	5	running	run	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	129	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	129	7	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	129	8	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	129	9	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	129	10	automatic	automatic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	129	11	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	129	12	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	129	13	can	can	MD
work_nvbfnacgujehpi52fk2nc7sg2m	129	14	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	129	15	run	run	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	129	16	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	129	17	times	time	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	129	18	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	129	19	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	129	20	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	129	21	increase	increase	VB
work_nvbfnacgujehpi52fk2nc7sg2m	129	22	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	129	23	reliabil-	reliabil-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	129	24	ity	ity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	129	25	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	129	26	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	129	27	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	129	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	130	1	Test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	2	specifications	specification	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	130	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	130	4	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	5	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	130	6	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	130	7	written	write	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	130	8	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	130	9	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	130	10	data	data	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	11	base	base	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	130	13	enabling	enable	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	130	14	fast	fast	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	130	15	creation	creation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	130	17	new	new	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	130	18	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	130	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	130	20	easy	easy	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	130	21	access	access	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	22	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	130	23	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	130	24	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	130	25	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	131	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	131	2	testing	testing	NN
work_nvbfnacgujehpi52fk2nc7sg2m	131	3	software	software	NN
work_nvbfnacgujehpi52fk2nc7sg2m	131	4	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	131	5	very	very	RB
work_nvbfnacgujehpi52fk2nc7sg2m	131	6	robust	robust	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	131	7	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	131	8	enables	enable	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	131	9	conducting	conduct	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	131	10	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	131	11	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	131	12	other	other	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	131	13	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	131	14	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	131	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	131	16	future	future	NN
work_nvbfnacgujehpi52fk2nc7sg2m	131	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	132	1	5.3	5.3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	132	2	Stability	stability	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	3	calibration	calibration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	4	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	132	5	process	process	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	6	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	132	7	data	datum	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	132	8	generation	generation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	9	includes	include	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	132	10	many	many	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	132	11	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	132	12	elements	element	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	132	13	such	such	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	132	14	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	132	15	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	16	generation	generation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	17	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	132	18	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	132	19	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	132	20	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	21	dictionary	dictionary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	132	22	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	132	23	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	24	selection	selection	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	25	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	132	26	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	132	27	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	132	28	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	29	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	132	30	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	132	31	degrading	degrade	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	132	32	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	132	33	docu-	docu-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	34	ment	ment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	132	35	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	133	1	A	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	133	2	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	133	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	133	4	predefined	predefine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	133	5	rules	rule	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	133	6	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	133	7	parameters	parameter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	133	8	con-	con-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	133	9	trol	trol	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	133	10	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	133	11	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	133	12	elements	element	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	133	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	133	14	yet	yet	CC
work_nvbfnacgujehpi52fk2nc7sg2m	133	15	they	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	133	16	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	133	17	some	some	DT
work_nvbfnacgujehpi52fk2nc7sg2m	133	18	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	133	19	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	133	20	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	133	21	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	133	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	134	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	2	larger	large	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	134	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	4	volume	volume	NN
work_nvbfnacgujehpi52fk2nc7sg2m	134	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	134	6	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	134	7	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	134	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	9	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	134	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	134	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	12	smaller	small	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	134	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	14	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	134	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	134	16	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	17	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	134	18	elements	element	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	134	19	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	134	20	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	134	21	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	134	22	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	134	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	135	1	Therefore	therefore	RB
work_nvbfnacgujehpi52fk2nc7sg2m	135	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	135	3	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	135	4	had	have	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	135	5	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	135	6	find	find	VB
work_nvbfnacgujehpi52fk2nc7sg2m	135	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	135	8	sufficient	sufficient	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	135	9	size	size	NN
work_nvbfnacgujehpi52fk2nc7sg2m	135	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	135	11	tested	test	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	135	12	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	135	13	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	135	14	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	135	15	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	135	16	reduce	reduce	VB
work_nvbfnacgujehpi52fk2nc7sg2m	135	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	135	18	random	random	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	135	19	factor	factor	NN
work_nvbfnacgujehpi52fk2nc7sg2m	135	20	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	135	21	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	135	22	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	135	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	136	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	136	2	ran	run	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	136	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	136	4	system	system	NN
work_nvbfnacgujehpi52fk2nc7sg2m	136	5	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	136	6	inputs	input	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	136	7	consisting	consist	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	136	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	136	9	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	136	10	growing	grow	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	136	11	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	136	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	136	13	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	136	14	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	136	15	plotted	plot	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	136	16	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	136	17	corresponding	corresponding	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	136	18	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	136	19	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	136	20	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	136	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	137	1	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	137	2	seen	see	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	137	3	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	137	4	Figure	figure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	137	5	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	137	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	137	7	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	137	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	137	9	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	137	10	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	137	11	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	137	12	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	137	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	137	14	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	137	15	increases	increase	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	137	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	137	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	137	18	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	137	19	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	137	20	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	137	21	stabilizes	stabilize	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	137	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	138	1	To	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	138	2	evaluate	evaluate	VB
work_nvbfnacgujehpi52fk2nc7sg2m	138	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	138	4	fluctuation	fluctuation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	138	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	138	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	138	7	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	138	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	138	9	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	138	10	calculate	calculate	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	138	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	138	12	standard	standard	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	138	13	deviation	deviation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	138	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	138	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	138	16	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	138	17	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	138	18	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	138	19	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	138	20	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	138	21	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	138	22	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	138	23	five	five	CD
work_nvbfnacgujehpi52fk2nc7sg2m	138	24	subsequent	subsequent	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	138	25	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	138	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	139	1	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	139	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	139	3	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	139	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	139	5	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	139	6	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	139	7	used	use	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	139	8	increases	increase	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	139	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	139	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	139	11	standard	standard	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	139	12	deviation	deviation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	139	13	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	139	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	139	15	subsequent	subsequent	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	139	16	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	139	17	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	139	18	decreases	decrease	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	139	19	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	140	1	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	140	2	evaluation	evaluation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	140	3	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	140	4	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	140	5	above	above	IN
work_nvbfnacgujehpi52fk2nc7sg2m	140	6	150	150	CD
work_nvbfnacgujehpi52fk2nc7sg2m	140	7	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	140	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	140	9	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	140	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	140	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	140	12	standard	standard	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	140	13	deviation	deviation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	140	14	stays	stay	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	140	15	below	below	IN
work_nvbfnacgujehpi52fk2nc7sg2m	140	16	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	140	17	%	%	NN
work_nvbfnacgujehpi52fk2nc7sg2m	140	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	141	1	According	accord	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	141	2	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	141	3	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	141	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	141	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	141	6	set	set	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	141	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	141	8	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	141	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	141	10	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	141	11	per	per	IN
work_nvbfnacgujehpi52fk2nc7sg2m	141	12	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	141	13	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	141	14	2000	2000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	141	15	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	141	16	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	141	17	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	141	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	141	19	i.e.	i.e.	FW
work_nvbfnacgujehpi52fk2nc7sg2m	142	1	over	over	IN
work_nvbfnacgujehpi52fk2nc7sg2m	142	2	200	200	CD
work_nvbfnacgujehpi52fk2nc7sg2m	142	3	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	142	4	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	142	5	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	142	6	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	142	7	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	142	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	143	1	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	143	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	144	1	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	144	2	Stability	Stability	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	144	3	calibration	calibration	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	4	graph	graph	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	5	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	144	6	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	7	grade	grade	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	8	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	144	9	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	144	10	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	144	11	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	144	12	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	144	13	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	144	14	function	function	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	144	16	line	line	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	17	count	count	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	18	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	144	19	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	144	20	Fig	Fig	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	144	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	145	1	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	145	2	Two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	145	3	examples	example	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	145	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	145	5	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	145	6	noised	noised	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	145	7	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	145	8	segments	segment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	145	9	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	145	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	145	11	conducted	conduct	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	145	12	experiments	experiment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	145	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	146	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	146	2	examples	example	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	146	3	use	use	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	146	4	dif-	dif-	CC
work_nvbfnacgujehpi52fk2nc7sg2m	146	5	ferent	ferent	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	146	6	degradation	degradation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	146	7	models	model	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	146	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	146	9	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	146	10	upper	upper	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	146	11	segment	segment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	146	12	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	146	13	degraded	degraded	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	146	14	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	146	15	Gaussian	Gaussian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	146	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	146	17	salt	salt	NN
work_nvbfnacgujehpi52fk2nc7sg2m	146	18	&	&	CC
work_nvbfnacgujehpi52fk2nc7sg2m	146	19	pepper	pepper	NN
work_nvbfnacgujehpi52fk2nc7sg2m	146	20	noise	noise	NN
work_nvbfnacgujehpi52fk2nc7sg2m	146	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	146	22	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	146	23	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	146	24	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	146	25	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	146	26	degraded	degrade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	146	27	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	146	28	Kanungo	Kanungo	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	146	29	’s	’s	POS
work_nvbfnacgujehpi52fk2nc7sg2m	146	30	model	model	NN
work_nvbfnacgujehpi52fk2nc7sg2m	146	31	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	147	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	147	2	same	same	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	147	3	degradation	degradation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	147	5	binarization	binarization	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	6	processes	process	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	7	were	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	147	8	applied	apply	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	147	9	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	10	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	147	11	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	147	12	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	13	Language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	14	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	147	16	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	17	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	18	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	19	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	147	20	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	21	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	22	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	147	23	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	147	24	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	147	25	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	26	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	147	27	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	28	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	29	en	en	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	30	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	31	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	32	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	33	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	34	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	147	35	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	36	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	37	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	38	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	39	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	40	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	41	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	42	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	43	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	147	44	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	147	45	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	147	46	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	47	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	48	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	49	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	147	50	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	51	<	<	XX
work_nvbfnacgujehpi52fk2nc7sg2m	147	52	br/	br/	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	147	53	>	>	XX
work_nvbfnacgujehpi52fk2nc7sg2m	147	54	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	147	55	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	147	56	5.4	5.4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	147	57	Experiments	experiment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	58	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	147	59	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	60	Next	next	RB
work_nvbfnacgujehpi52fk2nc7sg2m	147	61	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	147	62	describe	describe	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	147	63	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	147	64	experiments	experiment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	65	performed	perform	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	147	66	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	67	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	147	68	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	69	system	system	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	70	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	71	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	147	72	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	147	73	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	147	74	examined	examined	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	147	75	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	147	76	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	147	77	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	148	1	Average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	148	2	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	3	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	4	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	148	5	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	148	6	generated	generate	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	148	7	nine	nine	CD
work_nvbfnacgujehpi52fk2nc7sg2m	148	8	lan-	lan-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	9	guages	guage	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	148	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	148	11	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	148	12	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	148	13	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	148	14	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	15	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	16	average	average	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	17	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	148	18	be-	be-	XX
work_nvbfnacgujehpi52fk2nc7sg2m	148	19	tween	tween	NN
work_nvbfnacgujehpi52fk2nc7sg2m	148	20	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	148	21	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	148	22	10	10	CD
work_nvbfnacgujehpi52fk2nc7sg2m	148	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	149	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	149	2	standard	standard	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	149	3	deviation	deviation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	149	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	149	5	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	149	6	lengths	length	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	149	7	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	149	8	set	set	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	149	9	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	149	10	one	one	CD
work_nvbfnacgujehpi52fk2nc7sg2m	149	11	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	149	12	all	all	DT
work_nvbfnacgujehpi52fk2nc7sg2m	149	13	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	149	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	150	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	150	2	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	150	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	150	4	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	150	5	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	150	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	150	7	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	150	8	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	150	9	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	150	10	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	150	11	randomly	randomly	RB
work_nvbfnacgujehpi52fk2nc7sg2m	150	12	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	150	13	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	150	14	normal	normal	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	150	15	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	150	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	150	17	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	150	18	specified	specified	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	150	19	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	150	20	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	150	21	standard	standard	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	150	22	deviation	deviation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	150	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	151	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	151	2	experiment	experiment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	151	3	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	151	4	re-	re-	RB
work_nvbfnacgujehpi52fk2nc7sg2m	151	5	peated	peate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	151	6	three	three	CD
work_nvbfnacgujehpi52fk2nc7sg2m	151	7	times	time	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	151	8	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	151	9	validating	validate	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	151	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	151	11	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	151	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	152	1	Figure	figure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	152	2	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	3	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	152	4	two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	5	sample	sample	NN
work_nvbfnacgujehpi52fk2nc7sg2m	152	6	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	152	7	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	152	8	three	three	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	9	docu-	docu-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	152	10	ments	ment	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	152	11	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	152	12	three	three	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	13	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	152	14	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	152	15	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	152	16	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	152	17	aver-	aver-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	152	18	ages	age	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	152	19	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	152	20	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	152	22	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	23	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	152	24	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	152	25	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	152	26	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	152	27	respectively	respectively	RB
work_nvbfnacgujehpi52fk2nc7sg2m	152	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	153	1	Figure	figure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	153	2	6	6	CD
work_nvbfnacgujehpi52fk2nc7sg2m	153	3	summarizes	summarize	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	153	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	153	5	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	153	6	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	153	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	153	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	153	9	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	153	10	experiment	experiment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	153	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	154	1	It	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	154	2	displays	display	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	154	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	154	4	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	154	5	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	154	6	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	154	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	154	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	154	9	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	154	10	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	154	11	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	154	12	function	function	NN
work_nvbfnacgujehpi52fk2nc7sg2m	154	13	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	154	14	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	154	15	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	154	16	average	average	NN
work_nvbfnacgujehpi52fk2nc7sg2m	154	17	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	154	18	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	154	19	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	154	20	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	154	21	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	154	22	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	154	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	155	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	155	2	notice	notice	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	155	3	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	155	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	155	5	difference	difference	NN
work_nvbfnacgujehpi52fk2nc7sg2m	155	6	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	155	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	155	8	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	155	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	155	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	155	11	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	155	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	155	13	both	both	DT
work_nvbfnacgujehpi52fk2nc7sg2m	155	14	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	155	15	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	155	16	high	high	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	155	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	156	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	156	2	F-	F-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	156	3	Measure	measure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	156	4	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	156	5	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	156	6	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	156	7	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	156	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	156	9	short	short	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	156	10	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	156	11	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	156	12	very	very	RB
work_nvbfnacgujehpi52fk2nc7sg2m	156	13	low	low	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	156	14	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	156	15	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	156	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	156	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	156	18	above	above	IN
work_nvbfnacgujehpi52fk2nc7sg2m	156	19	80	80	CD
work_nvbfnacgujehpi52fk2nc7sg2m	156	20	%	%	NN
work_nvbfnacgujehpi52fk2nc7sg2m	156	21	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	156	22	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	156	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	157	1	As	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	157	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	157	3	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	157	4	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	5	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	6	increases	increase	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	157	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	157	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	157	9	F	f	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	10	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	157	11	measure	measure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	12	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	157	13	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	157	14	approaches	approach	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	157	15	100	100	CD
work_nvbfnacgujehpi52fk2nc7sg2m	157	16	%	%	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	17	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	157	18	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	157	19	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	157	20	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	157	21	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	157	22	decreases	decrease	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	157	23	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	157	24	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	157	25	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	157	26	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	157	27	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	28	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	157	29	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	157	30	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	157	31	above	above	RB
work_nvbfnacgujehpi52fk2nc7sg2m	157	32	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	158	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	158	2	believe	believe	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	158	3	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	4	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	158	5	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	158	6	because	because	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	7	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	158	8	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	158	9	fine-	fine-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	158	10	tuned	tune	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	158	11	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	12	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	158	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	158	14	where	where	WRB
work_nvbfnacgujehpi52fk2nc7sg2m	158	15	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	158	16	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	17	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	18	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	158	19	7	7	CD
work_nvbfnacgujehpi52fk2nc7sg2m	158	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	158	21	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	22	seen	see	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	158	23	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	24	Figure	figure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	25	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	158	26	Ascending	ascend	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	158	27	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	158	28	descending	descend	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	158	29	characters	character	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	158	30	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	158	31	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	158	32	experi-	experi-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	158	33	ment	ment	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	158	34	examines	examine	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	158	35	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	158	36	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	37	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	38	ascender	ascender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	39	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	158	40	des-	des-	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	158	41	cender	cender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	42	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	43	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	158	44	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	45	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	158	46	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	158	47	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	159	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	159	2	that	that	DT
work_nvbfnacgujehpi52fk2nc7sg2m	159	3	purpose	purpose	NN
work_nvbfnacgujehpi52fk2nc7sg2m	159	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	159	5	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	159	6	create	create	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	159	7	randomly	randomly	RB
work_nvbfnacgujehpi52fk2nc7sg2m	159	8	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	159	9	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	159	10	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	159	11	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	159	12	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	159	13	partial	partial	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	159	14	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	159	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	159	16	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	159	17	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	159	18	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	159	19	case	case	NN
work_nvbfnacgujehpi52fk2nc7sg2m	159	20	letters	letter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	159	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	160	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	160	2	alphabet	alphabet	NN
work_nvbfnacgujehpi52fk2nc7sg2m	160	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	160	4	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	160	5	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	160	6	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	160	7	consists	consist	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	160	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	160	9	14	14	CD
work_nvbfnacgujehpi52fk2nc7sg2m	160	10	characters	character	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	160	11	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	160	12	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	160	13	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	160	14	compos-	compos-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	160	15	ition	ition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	160	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	160	17	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	160	18	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	160	19	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	160	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	161	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	161	2	used	use	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	161	3	more	more	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	161	4	than	than	IN
work_nvbfnacgujehpi52fk2nc7sg2m	161	5	one	one	CD
work_nvbfnacgujehpi52fk2nc7sg2m	161	6	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	7	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	8	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	161	9	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	161	10	composition	composition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	11	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	161	12	eliminate	eliminate	VB
work_nvbfnacgujehpi52fk2nc7sg2m	161	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	161	14	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	161	16	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	161	17	choice	choice	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	18	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	161	19	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	161	20	specific	specific	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	161	21	character	character	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	22	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	161	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	162	1	Figure	figure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	2	7	7	CD
work_nvbfnacgujehpi52fk2nc7sg2m	162	3	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	162	4	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	162	5	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	162	6	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	162	7	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	162	8	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	162	9	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	10	each	each	DT
work_nvbfnacgujehpi52fk2nc7sg2m	162	11	combination	combination	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	12	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	162	13	#	#	$
work_nvbfnacgujehpi52fk2nc7sg2m	162	14	ascenders_#descenders	ascenders_#descenders	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	15	_	_	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	162	16	#	#	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	17	normal	normal	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	162	18	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	162	19	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	162	20	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	162	21	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	22	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	162	23	projection	projection	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	24	profile	profile	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	25	features	feature	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	162	26	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	162	27	significantly	significantly	RB
work_nvbfnacgujehpi52fk2nc7sg2m	162	28	better	well	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	162	29	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	162	30	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	31	balanced	balanced	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	162	32	com-	com-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	33	bination	bination	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	34	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	35	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	162	36	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	162	37	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	162	38	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	39	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	162	40	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	162	41	method	method	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	42	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	162	43	less	less	RBR
work_nvbfnacgujehpi52fk2nc7sg2m	162	44	influenced	influence	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	162	45	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	162	46	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	162	47	characteristic	characteristic	NN
work_nvbfnacgujehpi52fk2nc7sg2m	162	48	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	163	1	For	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	163	2	both	both	DT
work_nvbfnacgujehpi52fk2nc7sg2m	163	3	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	163	4	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	163	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	163	6	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	163	7	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	163	8	similar	similar	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	163	9	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	163	10	F	f	NN
work_nvbfnacgujehpi52fk2nc7sg2m	163	11	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	163	12	measure	measure	VB
work_nvbfnacgujehpi52fk2nc7sg2m	163	13	around	around	RB
work_nvbfnacgujehpi52fk2nc7sg2m	163	14	75	75	CD
work_nvbfnacgujehpi52fk2nc7sg2m	163	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	163	16	80	80	CD
work_nvbfnacgujehpi52fk2nc7sg2m	163	17	%	%	NN
work_nvbfnacgujehpi52fk2nc7sg2m	163	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	163	19	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	163	20	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	163	21	rec-	rec-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	163	22	ognition	ognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	163	23	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	163	24	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	163	25	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	163	26	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	163	27	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	163	28	no	no	DT
work_nvbfnacgujehpi52fk2nc7sg2m	163	29	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	163	30	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	163	31	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	163	32	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	163	33	much	much	RB
work_nvbfnacgujehpi52fk2nc7sg2m	163	34	lower	low	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	163	35	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	163	36	about	about	RB
work_nvbfnacgujehpi52fk2nc7sg2m	163	37	65.5	65.5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	163	38	%	%	NN
work_nvbfnacgujehpi52fk2nc7sg2m	163	39	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	163	40	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	164	1	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	164	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	165	1	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	165	2	Examples	Examples	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	165	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	165	4	spotting	spot	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	165	5	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	165	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	165	7	noisy	noisy	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	165	8	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	165	9	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	165	10	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	165	11	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	165	12	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	165	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	166	1	Two	two	CD
work_nvbfnacgujehpi52fk2nc7sg2m	166	2	example	example	NN
work_nvbfnacgujehpi52fk2nc7sg2m	166	3	lines	line	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	166	4	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	166	5	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	166	6	three	three	CD
work_nvbfnacgujehpi52fk2nc7sg2m	166	7	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	166	8	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	166	9	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	166	10	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	166	11	lengths	length	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	166	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	166	13	3	3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	166	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	166	15	5	5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	166	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	166	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	166	18	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	166	19	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	167	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	2	query	query	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	3	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	167	4	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	167	5	marked	mark	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	167	6	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	167	7	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	8	underline	underline	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	167	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	167	11	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	12	matches	match	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	167	13	found	find	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	167	14	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	167	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	16	system	system	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	17	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	167	18	marked	mark	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	167	19	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	167	20	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	21	rectangle	rectangle	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	22	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	167	23	therefore	therefore	RB
work_nvbfnacgujehpi52fk2nc7sg2m	167	24	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	25	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	26	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	167	27	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	28	first	first	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	167	29	line	line	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	30	marked	mark	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	167	31	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	167	32	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	33	rectangle	rectangle	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	34	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	167	35	not	not	RB
work_nvbfnacgujehpi52fk2nc7sg2m	167	36	underlined	underline	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	167	37	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	167	38	an	an	DT
work_nvbfnacgujehpi52fk2nc7sg2m	167	39	example	example	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	40	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	167	41	false	false	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	167	42	positive	positive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	167	43	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	167	44	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	167	45	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	168	1	7	7	CD
work_nvbfnacgujehpi52fk2nc7sg2m	168	2	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	168	3	graph	graph	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	4	depicts	depict	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	168	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	168	6	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	168	7	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	168	8	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	168	9	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	168	10	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	168	11	function	function	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	168	13	ascender	ascender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	14	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	168	15	descender	descender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	16	composition	composition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	17	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	168	18	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	168	19	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	20	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	168	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	169	1	6	6	CD
work_nvbfnacgujehpi52fk2nc7sg2m	169	2	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	3	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	4	experiment	experiment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	5	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	169	6	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	7	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	169	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	169	9	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	169	10	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	169	11	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	12	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	169	13	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	14	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	169	15	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	169	16	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	169	17	function	function	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	18	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	169	19	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	169	20	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	21	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	22	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	169	23	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	169	24	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	169	25	O.	O.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	26	Biller	Biller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	27	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	28	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	169	29	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	170	1	6	6	CD
work_nvbfnacgujehpi52fk2nc7sg2m	170	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	3	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	170	4	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	170	6	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	7	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	170	9	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	170	10	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	11	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	12	en	en	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	13	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	14	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	15	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	16	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	17	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	170	18	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	19	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	20	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	21	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	22	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	23	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	24	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	25	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	26	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	170	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	170	28	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	170	29	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	30	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	31	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	32	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	170	33	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	34	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	35	order	order	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	36	&	&	CC
work_nvbfnacgujehpi52fk2nc7sg2m	170	37	percnt	percnt	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	38	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	170	39	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	170	40	Orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	170	41	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	42	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	170	43	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	44	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	170	45	experiment	experiment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	46	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	170	47	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	170	48	created	create	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	170	49	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	170	50	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	51	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	170	52	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	170	53	neigh-	neigh-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	54	bor	bor	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	55	count	count	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	170	56	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	170	57	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	170	58	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	170	59	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	170	60	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	170	61	fixed	fix	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	170	62	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	170	63	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	64	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	170	65	to	to	TO
work_nvbfnacgujehpi52fk2nc7sg2m	170	66	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	170	67	7	7	CD
work_nvbfnacgujehpi52fk2nc7sg2m	170	68	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	170	69	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	171	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	171	2	created	create	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	171	3	test	test	NN
work_nvbfnacgujehpi52fk2nc7sg2m	171	4	set	set	NN
work_nvbfnacgujehpi52fk2nc7sg2m	171	5	contained	contain	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	171	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	171	7	values	value	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	171	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	171	9	0.5	0.5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	171	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	171	11	2	2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	171	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	171	13	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	171	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	171	15	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	171	16	6	6	CD
work_nvbfnacgujehpi52fk2nc7sg2m	171	17	as	as	RB
work_nvbfnacgujehpi52fk2nc7sg2m	171	18	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	171	19	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	171	20	neighbor	neighbor	NN
work_nvbfnacgujehpi52fk2nc7sg2m	171	21	count	count	NN
work_nvbfnacgujehpi52fk2nc7sg2m	171	22	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	172	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	172	2	graph	graph	NN
work_nvbfnacgujehpi52fk2nc7sg2m	172	3	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	4	Figure	Figure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	172	5	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	172	6	displays	display	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	172	7	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	172	8	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	172	9	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	172	10	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	172	11	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	172	12	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	13	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	172	14	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	172	15	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	172	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	172	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	172	18	shows	show	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	172	19	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	20	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	21	both	both	DT
work_nvbfnacgujehpi52fk2nc7sg2m	172	22	methods	method	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	172	23	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	172	24	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	172	25	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	26	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	172	27	declines	decline	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	172	28	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	29	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	172	30	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	172	31	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	172	32	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	172	33	neighboring	neighboring	NN
work_nvbfnacgujehpi52fk2nc7sg2m	172	34	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	172	35	grows	grow	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	172	36	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	173	1	6	6	CD
work_nvbfnacgujehpi52fk2nc7sg2m	173	2	Conclusions	Conclusions	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	173	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	173	4	Future	Future	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	173	5	Research	Research	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	173	6	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	173	7	this	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	173	8	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	173	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	173	10	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	173	11	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	173	12	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	173	13	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	173	14	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	173	15	several	several	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	173	16	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	173	17	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	173	18	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	173	19	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	173	20	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	173	21	rec-	rec-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	173	22	ognition	ognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	173	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	174	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	174	2	examine	examine	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	174	3	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	174	4	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	174	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	174	6	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	174	7	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	174	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	174	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	174	10	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	174	11	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	174	12	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	174	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	174	14	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	174	15	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	175	1	These	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	175	2	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	175	3	show	show	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	175	4	very	very	RB
work_nvbfnacgujehpi52fk2nc7sg2m	175	5	different	different	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	175	6	behaviour	behaviour	NN
work_nvbfnacgujehpi52fk2nc7sg2m	175	7	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	175	8	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	175	9	natural	natural	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	175	10	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	175	11	we	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	175	12	tested	test	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	175	13	;	;	:
work_nvbfnacgujehpi52fk2nc7sg2m	175	14	e.g.	e.g.	RB
work_nvbfnacgujehpi52fk2nc7sg2m	175	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	175	16	Hebrew	Hebrew	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	175	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	175	18	Arabic	Arabic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	175	19	have	have	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	175	20	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	175	21	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	175	22	shorter	short	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	175	23	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	175	24	than	than	IN
work_nvbfnacgujehpi52fk2nc7sg2m	175	25	English	English	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	175	26	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	175	27	Russian	Russian	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	175	28	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	175	29	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	175	30	larger	large	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	175	31	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	175	32	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	175	33	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	175	34	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	175	35	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	175	36	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	175	37	also	also	RB
work_nvbfnacgujehpi52fk2nc7sg2m	175	38	fewer	few	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	175	39	ascender	ascender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	175	40	–	–	:
work_nvbfnacgujehpi52fk2nc7sg2m	175	41	descender	descender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	175	42	configurations	configuration	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	175	43	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	176	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	176	2	investigated	investigate	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	176	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	176	4	influence	influence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	176	5	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	6	these	these	DT
work_nvbfnacgujehpi52fk2nc7sg2m	176	7	character-	character-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	176	8	istics	istic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	176	9	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	10	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	176	11	performance	performance	NN
work_nvbfnacgujehpi52fk2nc7sg2m	176	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	13	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	176	14	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	176	15	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	16	degraded	degraded	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	176	17	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	176	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	176	19	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	20	running	run	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	176	21	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	176	22	spotting	spot	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	176	23	tests	test	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	176	24	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	25	synthetically	synthetically	RB
work_nvbfnacgujehpi52fk2nc7sg2m	176	26	generated	generate	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	176	27	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	176	28	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	176	29	all	all	PDT
work_nvbfnacgujehpi52fk2nc7sg2m	176	30	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	176	31	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	176	32	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	176	33	were	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	176	34	constructed	construct	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	176	35	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	176	36	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	176	37	English	english	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	176	38	alphabet	alphabet	NN
work_nvbfnacgujehpi52fk2nc7sg2m	176	39	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	176	40	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	177	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	177	2	synthetic	synthetic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	177	3	languages	language	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	177	4	are	be	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	177	5	built	build	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	177	6	so	so	IN
work_nvbfnacgujehpi52fk2nc7sg2m	177	7	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	177	8	just	just	RB
work_nvbfnacgujehpi52fk2nc7sg2m	177	9	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	177	10	examined	examine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	177	11	characteristic	characteristic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	177	12	varies	varie	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	177	13	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	177	14	all	all	PDT
work_nvbfnacgujehpi52fk2nc7sg2m	177	15	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	177	16	others	other	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	177	17	remain	remain	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	177	18	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	177	19	same	same	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	177	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	178	1	We	-PRON-	PRP
work_nvbfnacgujehpi52fk2nc7sg2m	178	2	conclude	conclude	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	178	3	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	4	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	5	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	178	6	characteristics	characteristics	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	178	7	as-	as-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	178	8	cender	cender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	9	/	/	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	178	10	descender	descender	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	11	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	178	12	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	13	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	178	14	neighbor	neighbor	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	15	count	count	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	16	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	178	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	178	18	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	178	19	demonstrate	demonstrate	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	178	20	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	178	21	similar	similar	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	178	22	behavior	behavior	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	23	with	with	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	24	minor	minor	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	178	25	differences	difference	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	178	26	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	178	27	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	28	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	178	29	low	low	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	178	30	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	31	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	32	neighbors	neighbor	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	178	33	per	per	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	34	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	35	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	178	36	balanced	balanced	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	178	37	distribution	distribution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	38	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	39	ascenders	ascender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	178	40	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	178	41	descenders	descender	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	178	42	contribute	contribute	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	178	43	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	44	better	well	RBR
work_nvbfnacgujehpi52fk2nc7sg2m	178	45	re-	re-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	178	46	sults	sult	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	178	47	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	178	48	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	49	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	178	50	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	179	1	On	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	179	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	179	3	other	other	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	179	4	hand	hand	NN
work_nvbfnacgujehpi52fk2nc7sg2m	179	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	179	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	179	7	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	179	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	179	9	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	179	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	179	11	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	179	12	clearly	clearly	RB
work_nvbfnacgujehpi52fk2nc7sg2m	179	13	presents	present	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	179	14	better	well	JJR
work_nvbfnacgujehpi52fk2nc7sg2m	179	15	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	179	16	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	179	17	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	179	18	get	get	VBP
work_nvbfnacgujehpi52fk2nc7sg2m	179	19	longer	long	RBR
work_nvbfnacgujehpi52fk2nc7sg2m	179	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	179	21	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	179	22	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	179	23	responds	respond	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	179	24	best	best	RB
work_nvbfnacgujehpi52fk2nc7sg2m	179	25	when	when	WRB
work_nvbfnacgujehpi52fk2nc7sg2m	179	26	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	179	27	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	179	28	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	179	29	is	be	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	179	30	between	between	IN
work_nvbfnacgujehpi52fk2nc7sg2m	179	31	3–7	3–7	CD
work_nvbfnacgujehpi52fk2nc7sg2m	179	32	let-	let-	NN
work_nvbfnacgujehpi52fk2nc7sg2m	179	33	ters	ter	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	179	34	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	180	1	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	180	2	can	can	MD
work_nvbfnacgujehpi52fk2nc7sg2m	180	3	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	180	4	explained	explain	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	180	5	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	6	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	180	7	fact	fact	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	8	that	that	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	9	DTW	DTW	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	10	allows	allow	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	180	11	flexibility	flexibility	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	12	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	13	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	14	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	15	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	180	16	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	180	18	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	19	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	180	20	2007	2007	CD
work_nvbfnacgujehpi52fk2nc7sg2m	180	21	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	180	22	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	180	23	while	while	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	24	GSC	GSC	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	25	performs	perform	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	180	26	substitution	substitution	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	27	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	28	words	word	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	180	29	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	30	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	180	31	predefined	predefine	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	180	32	number	number	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	33	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	34	regions	region	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	180	35	within	within	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	36	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	180	37	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	38	frame	frame	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	39	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	180	40	Zhang	Zhang	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	41	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	42	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	43	.	.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	180	44	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	180	45	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	180	46	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	180	47	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	180	48	which	which	WDT
work_nvbfnacgujehpi52fk2nc7sg2m	180	49	leads	lead	VBZ
work_nvbfnacgujehpi52fk2nc7sg2m	180	50	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	51	optimal	optimal	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	180	52	behavior	behavior	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	53	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	54	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	180	55	limited	limited	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	180	56	range	range	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	57	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	180	58	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	180	59	lengths	length	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	180	60	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	181	1	Acknowledgements	acknowledgement	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	181	2	This	this	DT
work_nvbfnacgujehpi52fk2nc7sg2m	181	3	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	181	4	was	be	VBD
work_nvbfnacgujehpi52fk2nc7sg2m	181	5	supported	support	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	181	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	181	7	part	part	NN
work_nvbfnacgujehpi52fk2nc7sg2m	181	8	by	by	IN
work_nvbfnacgujehpi52fk2nc7sg2m	181	9	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	181	10	DFG-	DFG-	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	181	11	Trilateral	Trilateral	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	181	12	Grant	Grant	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	181	13	no	no	UH
work_nvbfnacgujehpi52fk2nc7sg2m	181	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	182	1	8716	8716	CD
work_nvbfnacgujehpi52fk2nc7sg2m	182	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	182	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	182	4	Lynn	Lynn	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	182	6	William	William	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	7	Frankel	Frankel	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	8	Center	Center	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	9	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	182	10	Computer	Computer	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	11	Sciences	Sciences	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	182	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	182	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	182	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	182	15	Paul	Paul	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	16	Ivanier	Ivanier	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	17	Center	Center	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	18	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	182	19	Robotics	Robotics	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	182	20	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	182	21	Production	Production	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	22	Management	Management	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	23	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	182	24	Ben	Ben	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	25	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	182	26	Gurion	Gurion	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	27	University	University	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	28	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	182	29	Israel	Israel	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	182	30	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	183	1	References	References	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	183	2	Andrews	Andrews	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	183	3	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	183	4	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	183	5	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	183	6	1989	1989	CD
work_nvbfnacgujehpi52fk2nc7sg2m	183	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	183	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	184	1	Frequency	frequency	NN
work_nvbfnacgujehpi52fk2nc7sg2m	184	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	184	3	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	184	4	effects	effect	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	184	5	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	184	6	lexical	lexical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	184	7	access	access	NN
work_nvbfnacgujehpi52fk2nc7sg2m	184	8	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	184	9	Activation	activation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	184	10	or	or	CC
work_nvbfnacgujehpi52fk2nc7sg2m	184	11	research	research	NN
work_nvbfnacgujehpi52fk2nc7sg2m	184	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	185	1	Journal	Journal	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	185	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	185	3	Experimental	Experimental	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	185	4	Psychology	Psychology	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	185	5	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	185	6	Learning	learning	NN
work_nvbfnacgujehpi52fk2nc7sg2m	185	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	185	8	Memory	memory	NN
work_nvbfnacgujehpi52fk2nc7sg2m	185	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	185	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	185	11	Cognition	Cognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	185	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	185	13	15	15	CD
work_nvbfnacgujehpi52fk2nc7sg2m	185	14	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	185	15	802–14	802–14	CD
work_nvbfnacgujehpi52fk2nc7sg2m	185	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	186	1	Andrews	Andrews	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	186	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	186	3	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	186	4	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	186	5	1992	1992	CD
work_nvbfnacgujehpi52fk2nc7sg2m	186	6	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	186	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	187	1	Frequency	frequency	NN
work_nvbfnacgujehpi52fk2nc7sg2m	187	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	187	3	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	187	4	effects	effect	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	187	5	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	187	6	lexical	lexical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	187	7	access	access	NN
work_nvbfnacgujehpi52fk2nc7sg2m	187	8	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	187	9	Lexical	lexical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	187	10	similarity	similarity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	187	11	or	or	CC
work_nvbfnacgujehpi52fk2nc7sg2m	187	12	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	187	13	re-	re-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	187	14	dundancy	dundancy	NN
work_nvbfnacgujehpi52fk2nc7sg2m	187	15	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	188	1	Journal	Journal	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	188	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	188	3	Experimental	Experimental	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	188	4	Psychology	Psychology	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	188	5	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	188	6	Learning	learning	NN
work_nvbfnacgujehpi52fk2nc7sg2m	188	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	188	8	Memory	memory	NN
work_nvbfnacgujehpi52fk2nc7sg2m	188	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	188	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	188	11	Cognition	Cognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	188	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	188	13	18	18	CD
work_nvbfnacgujehpi52fk2nc7sg2m	188	14	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	188	15	234–54	234–54	CD
work_nvbfnacgujehpi52fk2nc7sg2m	188	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	189	1	Andrews	Andrews	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	189	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	189	3	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	189	4	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	189	5	1997	1997	CD
work_nvbfnacgujehpi52fk2nc7sg2m	189	6	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	189	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	190	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	190	2	effects	effect	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	190	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	190	4	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	190	5	similarity	similarity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	190	6	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	190	7	lexical	lexical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	190	8	retrieval	retrieval	NN
work_nvbfnacgujehpi52fk2nc7sg2m	190	9	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	190	10	Resolving	resolve	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	190	11	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	190	12	conflicts	conflict	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	190	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	191	1	Psychological	Psychological	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	191	2	Bulletin	Bulletin	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	191	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	191	4	Review	Review	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	191	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	191	6	4	4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	191	7	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	191	8	439–61	439–61	CD
work_nvbfnacgujehpi52fk2nc7sg2m	191	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	192	1	Biadsy	Biadsy	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	192	3	F.	F.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	192	5	El	El	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	6	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	192	7	Sana	Sana	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	192	9	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	192	11	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	192	12	Habash	Habash	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	192	14	N.	N.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	192	15	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	192	16	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	192	17	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	192	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	193	1	Online	online	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	193	2	Arabic	arabic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	193	3	handwriting	handwriting	NN
work_nvbfnacgujehpi52fk2nc7sg2m	193	4	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	193	5	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	193	6	Hidden	Hidden	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	193	7	Markov	Markov	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	193	8	Models	Models	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	193	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	194	1	Proceedings	proceeding	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	194	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	194	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	194	4	10th	10th	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	194	5	International	International	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	6	Workshop	Workshop	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	7	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	194	8	Frontiers	Frontiers	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	194	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	194	10	Handwriting	Handwriting	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	11	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	194	12	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	194	14	La	La	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	15	Baule	Baule	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	194	17	Centre	Centre	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	18	de	de	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	19	Congres	Congres	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	20	Atlantia	Atlantia	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	21	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	194	22	France	France	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	194	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	195	1	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	195	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	196	1	1009–33	1009–33	CD
work_nvbfnacgujehpi52fk2nc7sg2m	196	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	197	1	Blando	Blando	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	197	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	197	3	L.	L.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	197	4	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	197	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	197	6	Kanai	Kanai	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	197	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	197	8	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	197	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	197	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	197	11	Nartker	Nartker	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	197	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	197	13	T.	t.	NN
work_nvbfnacgujehpi52fk2nc7sg2m	197	14	A.	a.	NN
work_nvbfnacgujehpi52fk2nc7sg2m	198	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	198	2	1995	1995	CD
work_nvbfnacgujehpi52fk2nc7sg2m	198	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	198	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	199	1	Prediction	prediction	NN
work_nvbfnacgujehpi52fk2nc7sg2m	199	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	199	3	OCR	OCR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	199	4	accuracy	accuracy	NN
work_nvbfnacgujehpi52fk2nc7sg2m	199	5	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	199	6	simple	simple	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	199	7	image	image	NN
work_nvbfnacgujehpi52fk2nc7sg2m	199	8	features	feature	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	199	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	200	1	Montreal	Montreal	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	200	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	200	3	Canada	Canada	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	200	4	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	200	5	ICDAR	ICDAR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	200	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	200	7	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	200	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	201	1	319–22	319–22	CD
work_nvbfnacgujehpi52fk2nc7sg2m	201	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	202	1	Fig	fig	NN
work_nvbfnacgujehpi52fk2nc7sg2m	202	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	203	1	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	2	F	F	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	3	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	203	4	Measure	Measure	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	5	results	result	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	203	6	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	203	7	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	203	8	function	function	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	203	10	average	average	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	203	11	neigh-	neigh-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	203	12	bor	bor	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	13	count	count	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	14	Language	Language	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	15	characteristics	characteristic	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	203	16	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	203	17	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	18	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	19	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	20	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	203	21	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	22	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	23	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	24	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	25	7	7	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	26	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	203	27	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	28	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	203	29	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	30	en	en	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	31	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	32	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	33	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	34	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	35	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	36	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	37	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	38	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	39	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	40	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	203	41	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	42	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	43	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	44	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	45	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	46	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	47	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	48	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	49	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	50	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	203	51	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	203	52	s	s	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	203	53	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	203	54	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	203	55	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	203	56	Coltheart	Coltheart	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	57	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	58	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	59	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	60	Davelaar	Davelaar	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	61	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	62	E.	E.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	63	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	64	Jonasson	Jonasson	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	65	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	66	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	67	F.	F.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	68	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	69	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	203	70	Besner	Besner	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	71	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	203	72	D.	D.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	203	73	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	203	74	1977	1977	CD
work_nvbfnacgujehpi52fk2nc7sg2m	203	75	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	203	76	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	204	1	Access	access	NN
work_nvbfnacgujehpi52fk2nc7sg2m	204	2	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	204	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	204	4	internal	internal	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	204	5	lexicon	lexicon	NN
work_nvbfnacgujehpi52fk2nc7sg2m	204	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	205	1	In	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	205	2	Dornic	Dornic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	205	3	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	205	4	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	206	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	206	2	ed	ed	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	206	3	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	207	1	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	207	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	207	3	Attention	attention	NN
work_nvbfnacgujehpi52fk2nc7sg2m	207	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	207	5	Performance	Performance	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	207	6	VI	VI	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	207	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	208	1	Hillsdale	Hillsdale	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	208	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	208	3	NJ	NJ	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	208	4	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	208	5	Erlbaum	Erlbaum	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	208	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	208	7	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	208	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	209	1	535–55	535–55	CD
work_nvbfnacgujehpi52fk2nc7sg2m	209	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	210	1	Esakov	Esakov	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	210	3	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	210	5	Lopresti	Lopresti	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	210	7	D.	D.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	8	P.	P.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	210	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	210	11	Sandberg	Sandberg	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	210	13	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	14	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	210	15	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	210	16	1994	1994	CD
work_nvbfnacgujehpi52fk2nc7sg2m	210	17	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	210	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	211	1	Classification	classification	NN
work_nvbfnacgujehpi52fk2nc7sg2m	211	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	211	3	Distribution	Distribution	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	211	5	Optical	Optical	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	6	Character	Character	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	7	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	8	Errors	Errors	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	211	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	211	10	Proceedings	Proceedings	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	211	11	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	211	12	SPIE	SPIE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	13	—	—	:
work_nvbfnacgujehpi52fk2nc7sg2m	211	14	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	211	15	International	International	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	16	Society	Society	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	17	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	211	18	Optical	Optical	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	19	Engineering	Engineering	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	211	21	Orlando	Orlando	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	22	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	211	23	Florida	Florida	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	211	25	2181	2181	CD
work_nvbfnacgujehpi52fk2nc7sg2m	211	26	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	211	27	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	211	28	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	212	1	204–16	204–16	CD
work_nvbfnacgujehpi52fk2nc7sg2m	212	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	213	1	Gatos	Gatos	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	3	B.	B.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	5	Konidaris	Konidaris	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	7	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	9	Ntzios	Ntzios	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	11	K.	K.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	13	Pratikakis	Pratikakis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	15	I.	I.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	16	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	17	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	213	18	Perantonis	Perantonis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	19	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	213	20	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	213	21	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	213	22	2005	2005	CD
work_nvbfnacgujehpi52fk2nc7sg2m	213	23	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	213	24	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	214	1	A	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	214	2	Segmentation	segmentation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	214	3	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	214	4	Free	Free	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	5	Approach	Approach	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	6	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	214	7	Keyword	Keyword	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	8	Search	Search	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	9	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	214	10	Historical	Historical	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	11	Typewritten	Typewritten	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	12	Documents	Documents	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	214	13	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	214	14	Proceedings	Proceedings	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	214	15	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	214	16	Eighth	Eighth	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	17	International	International	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	18	Conference	Conference	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	19	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	214	20	Document	Document	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	21	Analysis	Analysis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	22	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	214	23	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	214	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	214	25	Vol.1	vol.1	NN
work_nvbfnacgujehpi52fk2nc7sg2m	214	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	215	1	Seoul	Seoul	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	215	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	215	3	Korea	Korea	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	215	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	215	5	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	215	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	216	1	54–58	54–58	CD
work_nvbfnacgujehpi52fk2nc7sg2m	216	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	217	1	Grainger	Grainger	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	217	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	217	3	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	218	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	218	2	1990	1990	CD
work_nvbfnacgujehpi52fk2nc7sg2m	218	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	218	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	219	1	Word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	219	2	frequency	frequency	NN
work_nvbfnacgujehpi52fk2nc7sg2m	219	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	219	4	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	219	5	frequency	frequency	NN
work_nvbfnacgujehpi52fk2nc7sg2m	219	6	effects	effect	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	219	7	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	219	8	lexical	lexical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	219	9	decision	decision	NN
work_nvbfnacgujehpi52fk2nc7sg2m	219	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	219	11	naming	naming	NN
work_nvbfnacgujehpi52fk2nc7sg2m	219	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	220	1	Journal	Journal	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	220	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	220	3	Memory	Memory	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	220	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	220	5	Language	Language	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	220	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	220	7	29	29	CD
work_nvbfnacgujehpi52fk2nc7sg2m	220	8	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	220	9	228–44	228–44	CD
work_nvbfnacgujehpi52fk2nc7sg2m	220	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	221	1	Kanungo	Kanungo	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	221	3	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	221	5	Haralick	Haralick	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	221	7	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	8	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	221	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	221	11	Phillips	Phillips	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	221	13	I.	I.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	221	14	T.	t.	NN
work_nvbfnacgujehpi52fk2nc7sg2m	221	15	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	221	16	1993	1993	CD
work_nvbfnacgujehpi52fk2nc7sg2m	221	17	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	221	18	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	222	1	Global	global	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	222	2	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	222	3	Local	Local	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	4	Document	Document	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	5	Degradation	Degradation	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	6	Models	Models	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	222	7	ICDAR	ICDAR	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	222	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	222	9	Tsukuba	Tsukuba	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	10	City	City	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	11	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	222	12	Japan	Japan	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	13	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	222	14	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	222	16	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	222	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	223	1	730–4	730–4	CD
work_nvbfnacgujehpi52fk2nc7sg2m	223	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	224	1	Lavrenko	Lavrenko	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	224	3	V.	V.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	224	5	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	224	7	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	8	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	224	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	224	11	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	224	13	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	224	14	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	224	15	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	224	16	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	224	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	225	1	Holistic	Holistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	2	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	3	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	4	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	225	5	Handwritten	Handwritten	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	6	Historical	Historical	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	7	Documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	225	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	225	9	Document	Document	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	10	Image	Image	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	11	Analysis	Analysis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	225	12	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	225	13	Libraries	library	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	225	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	226	1	Palo	Palo	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	226	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	227	1	Alto	Alto	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	227	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	227	3	CA	CA	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	227	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	227	5	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	227	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	228	1	278–87	278–87	CD
work_nvbfnacgujehpi52fk2nc7sg2m	228	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	229	1	Levenshtein	Levenshtein	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	229	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	229	3	V.	V.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	229	4	I.	I.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	230	1	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	230	2	1966	1966	CD
work_nvbfnacgujehpi52fk2nc7sg2m	230	3	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	230	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	231	1	Binary	binary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	231	2	codes	code	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	231	3	capable	capable	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	231	4	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	231	5	cor-	cor-	RB
work_nvbfnacgujehpi52fk2nc7sg2m	231	6	recting	recting	NN
work_nvbfnacgujehpi52fk2nc7sg2m	231	7	deletions	deletion	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	231	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	231	9	insertions	insertion	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	231	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	231	11	reversals	reversal	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	231	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	232	1	Soviet	Soviet	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	232	2	Physics	Physics	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	232	3	Doklady	Doklady	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	232	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	232	5	10	10	CD
work_nvbfnacgujehpi52fk2nc7sg2m	232	6	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	232	7	707	707	CD
work_nvbfnacgujehpi52fk2nc7sg2m	232	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	233	1	Madhvanath	Madhvanath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	233	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	233	3	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	233	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	233	5	Govindaraju	Govindaraju	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	233	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	233	7	V.	V.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	233	8	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	233	9	2001	2001	CD
work_nvbfnacgujehpi52fk2nc7sg2m	233	10	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	233	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	234	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	234	2	role	role	NN
work_nvbfnacgujehpi52fk2nc7sg2m	234	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	234	4	holistic	holistic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	234	5	paradigms	paradigm	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	234	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	234	7	handwritten	handwritten	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	234	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	234	9	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	234	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	235	1	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	235	2	Transactions	Transactions	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	235	3	Pattern	Pattern	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	235	4	Analysis	Analysis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	235	5	Machine	Machine	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	235	6	Intelligence	Intelligence	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	235	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	235	8	23(2	23(2	CD
work_nvbfnacgujehpi52fk2nc7sg2m	235	9	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	235	10	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	235	11	149–64	149–64	CD
work_nvbfnacgujehpi52fk2nc7sg2m	235	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	236	1	New	New	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	3	B.	B.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	5	Ferrand	Ferrand	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	7	L.	L.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	9	Pallier	Pallier	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	11	C.	C.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	236	14	Brysbaert	Brysbaert	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	236	16	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	236	17	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	236	18	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	236	19	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	236	20	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	237	1	Reexamining	reexamine	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	237	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	237	3	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	4	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	5	effect	effect	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	6	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	237	7	visual	visual	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	237	8	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	9	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	10	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	237	11	New	new	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	237	12	evidence	evidence	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	13	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	237	14	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	237	15	english	english	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	237	16	lexi-	lexi-	XX
work_nvbfnacgujehpi52fk2nc7sg2m	237	17	con	con	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	237	18	project	project	NN
work_nvbfnacgujehpi52fk2nc7sg2m	237	19	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	238	1	Psychonomic	Psychonomic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	238	2	Bulletin	Bulletin	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	238	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	238	4	Review	Review	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	238	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	238	6	13(1	13(1	NN
work_nvbfnacgujehpi52fk2nc7sg2m	238	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	238	8	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	238	9	45–5	45–5	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	238	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	239	1	Perea	Perea	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	239	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	239	3	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	239	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	239	5	Rosa	Rosa	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	239	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	239	7	E.	E.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	239	8	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	239	9	2000	2000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	239	10	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	239	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	240	1	The	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	240	2	effects	effect	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	240	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	240	4	ortho-	ortho-	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	240	5	graphic	graphic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	240	6	neighborhood	neighborhood	NN
work_nvbfnacgujehpi52fk2nc7sg2m	240	7	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	240	8	reading	reading	NN
work_nvbfnacgujehpi52fk2nc7sg2m	240	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	240	10	laboratory	laboratory	NN
work_nvbfnacgujehpi52fk2nc7sg2m	240	11	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	240	12	identification	identification	NN
work_nvbfnacgujehpi52fk2nc7sg2m	240	13	tasks	task	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	240	14	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	240	15	A	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	240	16	review	review	NN
work_nvbfnacgujehpi52fk2nc7sg2m	240	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	241	1	Psicológica	Psicológica	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	241	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	241	3	21	21	CD
work_nvbfnacgujehpi52fk2nc7sg2m	241	4	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	241	5	327–40	327–40	CD
work_nvbfnacgujehpi52fk2nc7sg2m	241	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	242	1	Plamondon	Plamondon	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	242	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	242	3	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	242	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	242	5	Guerfali	Guerfali	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	242	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	242	7	W.	W.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	242	8	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	242	9	1996	1996	CD
work_nvbfnacgujehpi52fk2nc7sg2m	242	10	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	242	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	243	1	Why	why	WRB
work_nvbfnacgujehpi52fk2nc7sg2m	243	2	Handwriting	handwrite	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	243	3	Segmentation	segmentation	NN
work_nvbfnacgujehpi52fk2nc7sg2m	243	4	can	can	MD
work_nvbfnacgujehpi52fk2nc7sg2m	243	5	be	be	VB
work_nvbfnacgujehpi52fk2nc7sg2m	243	6	Misleading	mislead	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	243	7	?	?	.
work_nvbfnacgujehpi52fk2nc7sg2m	244	1	Proc	Proc	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	244	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	245	1	Intl	Intl	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	245	2	Conf	Conf	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	245	3	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	246	1	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	246	2	Pattern	Pattern	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	246	3	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	246	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	247	1	Vienna	Vienna	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	247	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	247	3	Austria	Austria	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	247	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	247	5	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	247	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	248	1	369–400	369–400	CD
work_nvbfnacgujehpi52fk2nc7sg2m	248	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	249	1	Plamondon	Plamondon	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	249	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	249	3	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	249	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	249	5	Srihari	Srihari	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	249	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	249	7	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	249	8	N.	N.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	249	9	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	249	10	2000	2000	CD
work_nvbfnacgujehpi52fk2nc7sg2m	249	11	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	249	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	250	1	On	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	250	2	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	250	3	line	line	NN
work_nvbfnacgujehpi52fk2nc7sg2m	250	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	250	5	off	off	IN
work_nvbfnacgujehpi52fk2nc7sg2m	250	6	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	250	7	line	line	NN
work_nvbfnacgujehpi52fk2nc7sg2m	250	8	handwriting	handwriting	NN
work_nvbfnacgujehpi52fk2nc7sg2m	250	9	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	250	10	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	250	11	A	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	250	12	comprehensive	comprehensive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	250	13	survey	survey	NN
work_nvbfnacgujehpi52fk2nc7sg2m	250	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	251	1	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	251	2	Transactions	Transactions	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	251	3	Pattern	Pattern	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	251	4	Analysis	Analysis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	251	5	Machine	Machine	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	251	6	Intelligence	Intelligence	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	251	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	251	8	22	22	CD
work_nvbfnacgujehpi52fk2nc7sg2m	251	9	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	251	10	63–84	63–84	CD
work_nvbfnacgujehpi52fk2nc7sg2m	251	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	252	1	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	252	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	252	3	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	252	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	252	5	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	252	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	252	7	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	252	8	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	252	9	2003a	2003a	CD
work_nvbfnacgujehpi52fk2nc7sg2m	252	10	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	252	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	253	1	Features	feature	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	253	2	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	253	3	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	253	4	Spotting	Spotting	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	253	5	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	253	6	Historical	historical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	253	7	Manuscripts	Manuscripts	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	253	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	254	1	Proceedings	Proceedings	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	254	3	Seventh	Seventh	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	4	International	International	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	5	Conference	Conference	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	6	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	254	7	Document	Document	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	8	Analysis	Analysis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	254	10	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	11	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	254	12	Vol	Vol	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	254	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	255	1	1	1	CD
work_nvbfnacgujehpi52fk2nc7sg2m	255	2	:	:	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	255	3	218–22	218–22	CD
work_nvbfnacgujehpi52fk2nc7sg2m	255	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	256	1	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	256	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	256	3	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	256	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	256	5	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	256	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	256	7	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	256	8	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	256	9	2003b	2003b	CD
work_nvbfnacgujehpi52fk2nc7sg2m	256	10	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	256	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	257	1	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	257	2	Image	Image	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	257	3	Matching	Matching	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	257	4	Using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	257	5	Dynamic	dynamic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	257	6	Time	time	NN
work_nvbfnacgujehpi52fk2nc7sg2m	257	7	Warping	Warping	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	257	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	258	1	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	2	Computer	Computer	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	3	Society	Society	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	4	Conference	Conference	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	5	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	258	6	Computer	Computer	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	7	Vision	Vision	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	8	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	258	9	Pattern	Pattern	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	10	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	11	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	258	12	CVPR’03	CVPR’03	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	13	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	258	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	258	15	Vol	Vol	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	258	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	259	1	2	2	LS
work_nvbfnacgujehpi52fk2nc7sg2m	259	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	259	3	II–521–7	ii–521–7	ADD
work_nvbfnacgujehpi52fk2nc7sg2m	259	4	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	260	1	Rath	Rath	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	260	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	260	3	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	260	4	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	260	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	260	6	Manmatha	Manmatha	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	260	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	260	8	R.	R.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	260	9	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	260	10	2007	2007	CD
work_nvbfnacgujehpi52fk2nc7sg2m	260	11	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	260	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	261	1	Word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	261	2	spotting	spot	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	261	3	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	261	4	historical	historical	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	261	5	documents	document	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	261	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	262	1	IJDAR	IJDAR	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	262	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	262	3	9(2–4	9(2–4	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	262	4	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	262	5	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	262	6	139–52	139–52	CD
work_nvbfnacgujehpi52fk2nc7sg2m	262	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	263	1	Richards	Richards	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	263	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	263	3	L.	L.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	263	4	G.	G.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	263	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	263	6	Heller	Heller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	263	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	263	8	F.	F.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	263	9	P.	P.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	263	10	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	263	11	1976	1976	CD
work_nvbfnacgujehpi52fk2nc7sg2m	263	12	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	263	13	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	264	1	Recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	264	2	thresholds	threshold	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	264	3	as	as	IN
work_nvbfnacgujehpi52fk2nc7sg2m	264	4	a	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	264	5	function	function	NN
work_nvbfnacgujehpi52fk2nc7sg2m	264	6	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	264	7	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	264	8	length	length	NN
work_nvbfnacgujehpi52fk2nc7sg2m	264	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	265	1	American	American	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	265	2	Journal	Journal	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	265	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	265	4	Psychology	Psychology	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	265	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	265	6	89(3	89(3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	265	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	265	8	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	265	9	455–66	455–66	CD
work_nvbfnacgujehpi52fk2nc7sg2m	265	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	266	1	Salah	Salah	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	266	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	266	3	A.	a.	NN
work_nvbfnacgujehpi52fk2nc7sg2m	267	1	B.	B.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	267	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	267	3	Ragot	Ragot	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	267	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	267	5	N.	N.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	267	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	267	7	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	267	8	Paquet	Paquet	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	267	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	267	10	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	267	11	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	267	12	2013	2013	CD
work_nvbfnacgujehpi52fk2nc7sg2m	267	13	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	267	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	268	1	Adaptive	adaptive	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	268	2	detection	detection	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	3	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	268	4	missed	miss	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	268	5	text	text	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	6	areas	area	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	268	7	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	268	8	OCR	ocr	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	9	outputs	output	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	268	10	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	268	11	Application	application	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	12	to	to	IN
work_nvbfnacgujehpi52fk2nc7sg2m	268	13	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	268	14	automatic	automatic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	268	15	assessment	assessment	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	16	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	268	17	OCR	OCR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	268	18	quality	quality	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	19	in	in	IN
work_nvbfnacgujehpi52fk2nc7sg2m	268	20	mass	mass	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	268	21	digitization	digitization	NN
work_nvbfnacgujehpi52fk2nc7sg2m	268	22	projects	project	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	268	23	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	269	1	IS&T	IS&T	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	269	2	/	/	SYM
work_nvbfnacgujehpi52fk2nc7sg2m	269	3	SPIE	SPIE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	269	4	Electronic	Electronic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	269	5	Imaging	Imaging	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	269	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	270	1	International	International	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	270	2	Society	Society	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	270	3	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	270	4	Optics	Optics	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	270	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	270	6	Photonics	Photonics	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	270	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	270	8	doi:10.1117/12.2003733	doi:10.1117/12.2003733	ADD
work_nvbfnacgujehpi52fk2nc7sg2m	270	9	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	271	1	Spitz	Spitz	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	271	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	271	3	A.	a.	NN
work_nvbfnacgujehpi52fk2nc7sg2m	271	4	L.	L.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	271	5	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	271	6	1997	1997	CD
work_nvbfnacgujehpi52fk2nc7sg2m	271	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	271	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	272	1	Determination	determination	NN
work_nvbfnacgujehpi52fk2nc7sg2m	272	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	272	3	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	272	4	script	script	NN
work_nvbfnacgujehpi52fk2nc7sg2m	272	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	272	6	language	language	NN
work_nvbfnacgujehpi52fk2nc7sg2m	272	7	content	content	NN
work_nvbfnacgujehpi52fk2nc7sg2m	272	8	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	272	9	document	document	NN
work_nvbfnacgujehpi52fk2nc7sg2m	272	10	images	image	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	272	11	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	273	1	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	273	2	Transactions	Transactions	NNPS
work_nvbfnacgujehpi52fk2nc7sg2m	273	3	Pattern	Pattern	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	273	4	Analysis	Analysis	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	273	5	Machine	Machine	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	273	6	Intelligence	Intelligence	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	273	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	273	8	19(3	19(3	CD
work_nvbfnacgujehpi52fk2nc7sg2m	273	9	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	273	10	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	273	11	235–45	235–45	CD
work_nvbfnacgujehpi52fk2nc7sg2m	273	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	274	1	Spitz	Spitz	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	274	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	274	3	A.	a.	NN
work_nvbfnacgujehpi52fk2nc7sg2m	274	4	L.	L.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	274	5	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	274	6	1999	1999	CD
work_nvbfnacgujehpi52fk2nc7sg2m	274	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	274	8	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	275	1	Shape	shape	NN
work_nvbfnacgujehpi52fk2nc7sg2m	275	2	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	275	3	based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	275	4	word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	275	5	recognition	recognition	NN
work_nvbfnacgujehpi52fk2nc7sg2m	275	6	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	276	1	ICDAR	ICDAR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	276	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	276	3	1(4	1(4	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	276	4	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	276	5	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	276	6	178–90	178–90	CD
work_nvbfnacgujehpi52fk2nc7sg2m	276	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	277	1	Yarkoni	Yarkoni	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	277	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	277	3	T.	T.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	277	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	277	5	Balota	Balota	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	277	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	277	7	D.	D.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	277	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	277	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	277	10	Yap	Yap	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	277	11	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	277	12	M.	M.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	277	13	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	277	14	2008	2008	CD
work_nvbfnacgujehpi52fk2nc7sg2m	277	15	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	277	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	278	1	Moving	move	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	278	2	beyond	beyond	IN
work_nvbfnacgujehpi52fk2nc7sg2m	278	3	colthearts	coltheart	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	278	4	n	n	XX
work_nvbfnacgujehpi52fk2nc7sg2m	278	5	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	278	6	A	a	DT
work_nvbfnacgujehpi52fk2nc7sg2m	278	7	new	new	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	278	8	measure	measure	NN
work_nvbfnacgujehpi52fk2nc7sg2m	278	9	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	278	10	orthographic	orthographic	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	278	11	similarity	similarity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	278	12	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	279	1	Psychonomic	Psychonomic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	279	2	Bulletin	Bulletin	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	279	3	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	279	4	Review	Review	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	279	5	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	279	6	15(5	15(5	CD
work_nvbfnacgujehpi52fk2nc7sg2m	279	7	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	279	8	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	279	9	971–9	971–9	CD
work_nvbfnacgujehpi52fk2nc7sg2m	279	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	280	1	Ye	Ye	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	3	P.	P.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	5	Kumar	Kumar	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	7	J.	J.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	9	Kang	Kang	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	11	L.	L.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	13	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	280	14	Doermann	Doermann	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	15	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	280	16	D.	D.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	17	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	280	18	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	280	19	2012	2012	CD
work_nvbfnacgujehpi52fk2nc7sg2m	280	20	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	280	21	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	281	1	Unsupervised	unsupervised	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	281	2	Feature	Feature	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	3	Learning	Learning	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	4	Framework	Framework	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	5	for	for	IN
work_nvbfnacgujehpi52fk2nc7sg2m	281	6	No	no	DT
work_nvbfnacgujehpi52fk2nc7sg2m	281	7	-	-	HYPH
work_nvbfnacgujehpi52fk2nc7sg2m	281	8	Reference	Reference	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	9	Image	Image	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	10	Quality	Quality	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	11	Assessment	Assessment	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	281	13	CVPR	CVPR	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	281	14	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	282	1	Providence	Providence	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	282	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	282	3	RI	RI	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	282	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	282	5	USA	USA	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	282	6	:	:	:
work_nvbfnacgujehpi52fk2nc7sg2m	282	7	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	282	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	282	9	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	282	10	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	283	1	1098–105	1098–105	CD
work_nvbfnacgujehpi52fk2nc7sg2m	283	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	284	1	Zagoris	Zagoris	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	284	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	284	3	K.	K.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	284	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	284	5	Papamarkos	Papamarkos	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	284	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	284	7	N.	N.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	284	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	284	9	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	284	10	Chamzas	Chamzas	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	284	11	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	284	12	C.	C.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	284	13	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	284	14	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	284	15	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	284	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	285	1	Web	Web	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	2	Document	Document	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	3	Image	Image	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	4	Retrieval	Retrieval	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	5	System	System	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	6	Based	base	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	285	7	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	285	8	Word	Word	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	9	Spotting	Spotting	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	10	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	285	11	IEEE	IEEE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	12	International	International	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	13	Conference	Conference	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	14	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	285	15	Image	Image	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	16	Processing	Processing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	17	2006	2006	CD
work_nvbfnacgujehpi52fk2nc7sg2m	285	18	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	285	19	Atlanta	Atlanta	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	20	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	285	21	Georgia	Georgia	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	22	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	285	23	USA	USA	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	24	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	285	25	pp	pp	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	285	26	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	286	1	477–80	477–80	CD
work_nvbfnacgujehpi52fk2nc7sg2m	286	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	287	1	Zhang	Zhang	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	2	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	287	3	B.	B.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	4	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	287	5	Srihari	Srihari	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	6	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	287	7	S.	S.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	8	N.	N.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	287	10	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	287	11	Huang	Huang	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	287	13	C.	C.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	287	14	(	(	-LRB-
work_nvbfnacgujehpi52fk2nc7sg2m	287	15	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	287	16	)	)	-RRB-
work_nvbfnacgujehpi52fk2nc7sg2m	287	17	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	288	1	Word	word	NN
work_nvbfnacgujehpi52fk2nc7sg2m	288	2	image	image	NN
work_nvbfnacgujehpi52fk2nc7sg2m	288	3	retrieval	retrieval	NN
work_nvbfnacgujehpi52fk2nc7sg2m	288	4	using	use	VBG
work_nvbfnacgujehpi52fk2nc7sg2m	288	5	binary	binary	JJ
work_nvbfnacgujehpi52fk2nc7sg2m	288	6	features	feature	NNS
work_nvbfnacgujehpi52fk2nc7sg2m	288	7	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	289	1	Proc	Proc	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	289	2	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	290	1	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	290	2	the	the	DT
work_nvbfnacgujehpi52fk2nc7sg2m	290	3	SPIE	SPIE	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	290	4	Conf	Conf	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	290	5	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	291	1	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	291	2	Document	Document	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	3	Recognition	Recognition	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	4	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	291	5	Retrieval	Retrieval	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	6	XI	XI	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	7	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	291	8	2004	2004	CD
work_nvbfnacgujehpi52fk2nc7sg2m	291	9	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	291	10	San	San	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	11	Jose	Jose	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	12	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	291	13	California	California	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	14	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	291	15	USA	USA	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	291	16	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	292	1	O.	O.	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	292	2	Biller	Biller	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	292	3	et	et	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	292	4	al	al	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	292	5	.	.	.
work_nvbfnacgujehpi52fk2nc7sg2m	293	1	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	293	2	of	of	IN
work_nvbfnacgujehpi52fk2nc7sg2m	293	3	8	8	CD
work_nvbfnacgujehpi52fk2nc7sg2m	293	4	Literary	Literary	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	5	and	and	CC
work_nvbfnacgujehpi52fk2nc7sg2m	293	6	Linguistic	Linguistic	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	7	Computing	Computing	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	8	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	293	9	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	293	10	at	at	IN
work_nvbfnacgujehpi52fk2nc7sg2m	293	11	B	b	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	12	en	en	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	13	G	G	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	14	urion	urion	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	15	U	U	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	16	niversity	niversity	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	17	-	-	,
work_nvbfnacgujehpi52fk2nc7sg2m	293	18	A	a	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	19	ranne	ranne	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	20	L	l	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	21	ibrary	ibrary	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	22	on	on	IN
work_nvbfnacgujehpi52fk2nc7sg2m	293	23	N	N	NNP
work_nvbfnacgujehpi52fk2nc7sg2m	293	24	ovem	ovem	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	25	ber	ber	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	26	20	20	CD
work_nvbfnacgujehpi52fk2nc7sg2m	293	27	,	,	,
work_nvbfnacgujehpi52fk2nc7sg2m	293	28	2014	2014	CD
work_nvbfnacgujehpi52fk2nc7sg2m	293	29	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	30	D	d	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	31	ow	ow	NN
work_nvbfnacgujehpi52fk2nc7sg2m	293	32	nloaded	nloade	VBN
work_nvbfnacgujehpi52fk2nc7sg2m	293	33	from	from	IN
work_nvbfnacgujehpi52fk2nc7sg2m	293	34	http://llc.oxfordjournals.org/	http://llc.oxfordjournals.org/	NN