id	sid	tid	token	lemma	pos
cril-249	1	1	language	language	NOUN
cril-249	1	2	identification	identification	NOUN
cril-249	1	3	and	and	CCONJ
cril-249	1	4	language	language	NOUN
cril-249	1	5	specific	specific	ADJ
cril-249	1	6	letter	letter	NOUN
cril-249	1	7	-	-	PUNCT
cril-249	1	8	to	to	ADP
cril-249	1	9	-	-	PUNCT
cril-249	1	10	sound	sound	NOUN
cril-249	1	11	rules	rule	NOUN
cril-249	1	12	language	language	NOUN
cril-249	1	13	identification	identification	NOUN
cril-249	1	14	and	and	CCONJ
cril-249	1	15	language	language	NOUN
cril-249	1	16	specific	specific	ADJ
cril-249	1	17	letter	letter	NOUN
cril-249	1	18	-	-	PUNCT
cril-249	1	19	to	to	ADP
cril-249	1	20	-	-	PUNCT
cril-249	1	21	sound	sound	NOUN
cril-249	1	22	rules	rule	NOUN
cril-249	1	23	stephen	stephen	PROPN
cril-249	1	24	lewis	lewis	PROPN
cril-249	1	25	,	,	PUNCT
cril-249	1	26	katie	katie	PROPN
cril-249	1	27	mcgrath	mcgrath	PROPN
cril-249	1	28	,	,	PUNCT
cril-249	1	29	jeffrey	jeffrey	PROPN
cril-249	1	30	reuppel	reuppel	PROPN
cril-249	1	31	university	university	PROPN
cril-249	1	32	of	of	ADP
cril-249	1	33	colorado	colorado	PROPN
cril-249	1	34	at	at	ADP
cril-249	1	35	boulder	boulder	NOUN
cril-249	1	36	this	this	DET
cril-249	1	37	paper	paper	NOUN
cril-249	1	38	describes	describe	VERB
cril-249	1	39	a	a	DET
cril-249	1	40	system	system	NOUN
cril-249	1	41	that	that	PRON
cril-249	1	42	improves	improve	VERB
cril-249	1	43	automatic	automatic	ADJ
cril-249	1	44	arpabet	arpabet	ADJ
cril-249	1	45	transcription	transcription	NOUN
cril-249	1	46	by	by	ADP
cril-249	1	47	addressing	address	VERB
cril-249	1	48	performance	performance	NOUN
cril-249	1	49	issues	issue	NOUN
cril-249	1	50	resulting	result	VERB
cril-249	1	51	from	from	ADP
cril-249	1	52	arabic	arabic	ADJ
cril-249	1	53	and	and	CCONJ
cril-249	1	54	russian	russian	ADJ
cril-249	1	55	transliteration	transliteration	NOUN
cril-249	1	56	in	in	ADP
cril-249	1	57	english	english	ADJ
cril-249	1	58	text	text	NOUN
cril-249	1	59	.	.	PUNCT
cril-249	2	1	our	our	PRON
cril-249	2	2	system	system	NOUN
cril-249	2	3	is	be	AUX
cril-249	2	4	called	call	VERB
cril-249	2	5	ear	ear	NOUN
cril-249	2	6	(	(	PUNCT
cril-249	2	7	english	english	PROPN
cril-249	2	8	,	,	PUNCT
cril-249	2	9	arabic	arabic	ADJ
cril-249	2	10	,	,	PUNCT
cril-249	2	11	russian	russian	PROPN
cril-249	2	12	)	)	PUNCT
cril-249	2	13	.	.	PUNCT
cril-249	3	1	the	the	DET
cril-249	3	2	ear	ear	NOUN
cril-249	3	3	system	system	NOUN
cril-249	3	4	has	have	VERB
cril-249	3	5	two	two	NUM
cril-249	3	6	components	component	NOUN
cril-249	3	7	:	:	PUNCT
cril-249	3	8	1	1	X
cril-249	3	9	.	.	X
cril-249	3	10	an	an	DET
cril-249	3	11	n	n	CCONJ
cril-249	3	12	-	-	PUNCT
cril-249	3	13	gram	gram	NOUN
cril-249	3	14	language	language	NOUN
cril-249	3	15	identifier	identifier	NOUN
cril-249	3	16	module	module	NOUN
cril-249	3	17	which	which	PRON
cril-249	3	18	classifies	classify	VERB
cril-249	3	19	an	an	DET
cril-249	3	20	incoming	incoming	ADJ
cril-249	3	21	unknown	unknown	ADJ
cril-249	3	22	word	word	NOUN
cril-249	3	23	as	as	ADP
cril-249	3	24	arabic	arabic	ADJ
cril-249	3	25	,	,	PUNCT
cril-249	3	26	russian	russian	ADJ
cril-249	3	27	,	,	PUNCT
cril-249	3	28	or	or	CCONJ
cril-249	3	29	english	english	PROPN
cril-249	3	30	,	,	PUNCT
cril-249	3	31	2	2	NUM
cril-249	3	32	.	.	X
cril-249	3	33	language	language	NOUN
cril-249	3	34	specific	specific	ADJ
cril-249	3	35	letter	letter	NOUN
cril-249	3	36	to	to	PART
cril-249	3	37	sound	sound	VERB
cril-249	3	38	rules	rule	NOUN
cril-249	3	39	which	which	PRON
cril-249	3	40	output	output	VERB
cril-249	3	41	a	a	DET
cril-249	3	42	pronunciation	pronunciation	NOUN
cril-249	3	43	for	for	ADP
cril-249	3	44	a	a	DET
cril-249	3	45	word	word	NOUN
cril-249	3	46	based	base	VERB
cril-249	3	47	on	on	ADP
cril-249	3	48	its	its	PRON
cril-249	3	49	classification	classification	NOUN
cril-249	3	50	.	.	PUNCT
cril-249	4	1	our	our	PRON
cril-249	4	2	results	result	NOUN
cril-249	4	3	show	show	VERB
cril-249	4	4	overall	overall	ADJ
cril-249	4	5	system	system	NOUN
cril-249	4	6	error	error	NOUN
cril-249	4	7	reduction	reduction	NOUN
cril-249	4	8	rates	rate	NOUN
cril-249	4	9	at	at	ADP
cril-249	4	10	upwards	upwards	ADV
cril-249	4	11	of	of	ADP
cril-249	4	12	45	45	NUM
cril-249	4	13	%	%	NOUN
cril-249	4	14	as	as	ADP
cril-249	4	15	compared	compare	VERB
cril-249	4	16	to	to	ADP
cril-249	4	17	a	a	DET
cril-249	4	18	system	system	NOUN
cril-249	4	19	trained	train	VERB
cril-249	4	20	only	only	ADV
cril-249	4	21	on	on	ADP
cril-249	4	22	english	english	PROPN
cril-249	4	23	.	.	PUNCT
cril-249	5	1	1	1	X
cril-249	5	2	.	.	X
cril-249	5	3	introduction	introduction	NOUN
cril-249	5	4	the	the	DET
cril-249	5	5	sparsity	sparsity	NOUN
cril-249	5	6	of	of	ADP
cril-249	5	7	transcribed	transcribe	VERB
cril-249	5	8	conversational	conversational	ADJ
cril-249	5	9	english	english	PROPN
cril-249	5	10	makes	make	VERB
cril-249	5	11	assembling	assemble	VERB
cril-249	5	12	a	a	DET
cril-249	5	13	corpus	corpus	NOUN
cril-249	5	14	for	for	ADP
cril-249	5	15	speech	speech	NOUN
cril-249	5	16	recognition	recognition	NOUN
cril-249	5	17	training	train	VERB
cril-249	5	18	a	a	DET
cril-249	5	19	challenging	challenging	ADJ
cril-249	5	20	task	task	NOUN
cril-249	5	21	.	.	PUNCT
cril-249	6	1	one	one	NUM
cril-249	6	2	alternative	alternative	ADJ
cril-249	6	3	resource	resource	NOUN
cril-249	6	4	for	for	ADP
cril-249	6	5	dealing	deal	VERB
cril-249	6	6	with	with	ADP
cril-249	6	7	sparsity	sparsity	NOUN
cril-249	6	8	is	be	AUX
cril-249	6	9	to	to	PART
cril-249	6	10	mine	mine	VERB
cril-249	6	11	the	the	DET
cril-249	6	12	world	world	NOUN
cril-249	6	13	wide	wide	ADJ
cril-249	6	14	web	web	NOUN
cril-249	6	15	for	for	ADP
cril-249	6	16	transcribed	transcribe	VERB
cril-249	6	17	conversations	conversation	NOUN
cril-249	6	18	.	.	PUNCT
cril-249	7	1	utilizing	utilize	VERB
cril-249	7	2	this	this	DET
cril-249	7	3	resource	resource	NOUN
cril-249	7	4	though	though	SCONJ
cril-249	7	5	poses	pose	VERB
cril-249	7	6	a	a	DET
cril-249	7	7	number	number	NOUN
cril-249	7	8	of	of	ADP
cril-249	7	9	new	new	ADJ
cril-249	7	10	problems	problem	NOUN
cril-249	7	11	from	from	ADP
cril-249	7	12	text	text	NOUN
cril-249	7	13	normalization	normalization	NOUN
cril-249	7	14	to	to	ADP
cril-249	7	15	producing	produce	VERB
cril-249	7	16	optimized	optimize	VERB
cril-249	7	17	output	output	NOUN
cril-249	7	18	for	for	ADP
cril-249	7	19	letter	letter	NOUN
cril-249	7	20	to	to	ADP
cril-249	7	21	speech	speech	NOUN
cril-249	7	22	.	.	PUNCT
cril-249	8	1	this	this	DET
cril-249	8	2	transcribed	transcribe	VERB
cril-249	8	3	english	english	ADJ
cril-249	8	4	text	text	NOUN
cril-249	8	5	,	,	PUNCT
cril-249	8	6	especially	especially	ADV
cril-249	8	7	news	news	NOUN
cril-249	8	8	text	text	NOUN
cril-249	8	9	,	,	PUNCT
cril-249	8	10	contains	contain	VERB
cril-249	8	11	significant	significant	ADJ
cril-249	8	12	a	a	DET
cril-249	8	13	number	number	NOUN
cril-249	8	14	of	of	ADP
cril-249	8	15	transliterated	transliterated	ADJ
cril-249	8	16	foreign	foreign	ADJ
cril-249	8	17	proper	proper	ADJ
cril-249	8	18	nouns	noun	NOUN
cril-249	8	19	.	.	PUNCT
cril-249	9	1	without	without	ADP
cril-249	9	2	normalization	normalization	NOUN
cril-249	9	3	the	the	DET
cril-249	9	4	output	output	NOUN
cril-249	9	5	of	of	ADP
cril-249	9	6	a	a	DET
cril-249	9	7	text	text	NOUN
cril-249	9	8	-	-	PUNCT
cril-249	9	9	to	to	ADP
cril-249	9	10	-	-	PUNCT
cril-249	9	11	speech	speech	NOUN
cril-249	9	12	system	system	NOUN
cril-249	9	13	will	will	AUX
cril-249	9	14	often	often	ADV
cril-249	9	15	generate	generate	VERB
cril-249	9	16	inappropriate	inappropriate	ADJ
cril-249	9	17	pronunciations	pronunciation	NOUN
cril-249	9	18	because	because	SCONJ
cril-249	9	19	the	the	DET
cril-249	9	20	letter	letter	NOUN
cril-249	9	21	-	-	PUNCT
cril-249	9	22	to	to	ADP
cril-249	9	23	-	-	PUNCT
cril-249	9	24	sound	sound	NOUN
cril-249	9	25	rules	rule	NOUN
cril-249	9	26	have	have	AUX
cril-249	9	27	been	be	AUX
cril-249	9	28	trained	train	VERB
cril-249	9	29	on	on	ADP
cril-249	9	30	english	english	ADJ
cril-249	9	31	spelling	spelling	NOUN
cril-249	9	32	standards	standard	NOUN
cril-249	9	33	,	,	PUNCT
cril-249	9	34	not	not	PART
cril-249	9	35	on	on	ADP
cril-249	9	36	foreign	foreign	ADJ
cril-249	9	37	transliteration	transliteration	NOUN
cril-249	9	38	standards	standard	NOUN
cril-249	9	39	.	.	PUNCT
cril-249	10	1	applying	apply	VERB
cril-249	10	2	english	english	ADJ
cril-249	10	3	letter	letter	NOUN
cril-249	10	4	-	-	PUNCT
cril-249	10	5	to	to	ADP
cril-249	10	6	-	-	PUNCT
cril-249	10	7	sound	sound	NOUN
cril-249	10	8	rules	rule	NOUN
cril-249	10	9	to	to	ADP
cril-249	10	10	arabic	arabic	ADJ
cril-249	10	11	transliteration	transliteration	NOUN
cril-249	10	12	,	,	PUNCT
cril-249	10	13	for	for	ADP
cril-249	10	14	instance	instance	NOUN
cril-249	10	15	,	,	PUNCT
cril-249	10	16	can	can	AUX
cril-249	10	17	result	result	VERB
cril-249	10	18	in	in	ADP
cril-249	10	19	the	the	DET
cril-249	10	20	following	follow	VERB
cril-249	10	21	mispronunciation	mispronunciation	NOUN
cril-249	10	22	.	.	PUNCT
cril-249	11	1	example	example	NOUN
cril-249	12	1	1	1	NUM
cril-249	12	2	.	.	X
cril-249	13	1	hibaaq	hibaaq	NOUN
cril-249	13	2	hh	hh	PROPN
cril-249	13	3	ay	ay	PROPN
cril-249	13	4	b	b	PROPN
cril-249	13	5	ae	ae	PROPN
cril-249	13	6	kd1	kd1	PROPN
cril-249	13	7	to	to	PART
cril-249	13	8	address	address	VERB
cril-249	13	9	the	the	DET
cril-249	13	10	issue	issue	NOUN
cril-249	13	11	of	of	ADP
cril-249	13	12	poorly	poorly	ADV
cril-249	13	13	produced	produce	VERB
cril-249	13	14	pronunciation	pronunciation	NOUN
cril-249	13	15	data	datum	NOUN
cril-249	13	16	due	due	ADP
cril-249	13	17	to	to	ADP
cril-249	13	18	the	the	DET
cril-249	13	19	presence	presence	NOUN
cril-249	13	20	of	of	ADP
cril-249	13	21	non	non	ADJ
cril-249	13	22	-	-	ADJ
cril-249	13	23	english	english	ADJ
cril-249	13	24	words	word	NOUN
cril-249	13	25	,	,	PUNCT
cril-249	13	26	we	we	PRON
cril-249	13	27	have	have	AUX
cril-249	13	28	implemented	implement	VERB
cril-249	13	29	a	a	DET
cril-249	13	30	two	two	NUM
cril-249	13	31	-	-	PUNCT
cril-249	13	32	tiered	tiere	VERB
cril-249	13	33	language	language	NOUN
cril-249	13	34	classifier	classifier	NOUN
cril-249	13	35	.	.	PUNCT
cril-249	14	1	the	the	DET
cril-249	14	2	first	first	ADJ
cril-249	14	3	tier	tier	NOUN
cril-249	14	4	of	of	ADP
cril-249	14	5	this	this	DET
cril-249	14	6	classifier	classifier	NOUN
cril-249	14	7	serves	serve	VERB
cril-249	14	8	to	to	PART
cril-249	14	9	identify	identify	VERB
cril-249	14	10	non	non	ADJ
cril-249	14	11	-	-	ADJ
cril-249	14	12	english	english	ADJ
cril-249	14	13	words	word	NOUN
cril-249	14	14	in	in	ADP
cril-249	14	15	the	the	DET
cril-249	14	16	conversational	conversational	ADJ
cril-249	14	17	data	datum	NOUN
cril-249	14	18	scraped	scrape	VERB
cril-249	14	19	from	from	ADP
cril-249	14	20	the	the	DET
cril-249	14	21	web	web	NOUN
cril-249	14	22	.	.	PUNCT
cril-249	15	1	the	the	DET
cril-249	15	2	second	second	ADJ
cril-249	15	3	tier	tier	NOUN
cril-249	15	4	processes	process	VERB
cril-249	15	5	the	the	DET
cril-249	15	6	results	result	NOUN
cril-249	15	7	of	of	ADP
cril-249	15	8	this	this	DET
cril-249	15	9	classification	classification	NOUN
cril-249	15	10	through	through	ADP
cril-249	15	11	specific	specific	ADJ
cril-249	15	12	letter	letter	NOUN
cril-249	15	13	-	-	PUNCT
cril-249	15	14	to	to	ADP
cril-249	15	15	-	-	PUNCT
cril-249	15	16	sound	sound	NOUN
cril-249	15	17	rules	rule	NOUN
cril-249	15	18	that	that	PRON
cril-249	15	19	have	have	AUX
cril-249	15	20	been	be	AUX
cril-249	15	21	trained	train	VERB
cril-249	15	22	for	for	ADP
cril-249	15	23	each	each	DET
cril-249	15	24	non	non	ADJ
cril-249	15	25	-	-	ADJ
cril-249	15	26	english	english	ADJ
cril-249	15	27	language	language	NOUN
cril-249	15	28	in	in	ADP
cril-249	15	29	question	question	NOUN
cril-249	15	30	.	.	PUNCT
cril-249	16	1	more	more	ADV
cril-249	16	2	specifically	specifically	ADV
cril-249	16	3	,	,	PUNCT
cril-249	16	4	our	our	PRON
cril-249	16	5	research	research	NOUN
cril-249	16	6	has	have	AUX
cril-249	16	7	focused	focus	VERB
cril-249	16	8	on	on	ADP
cril-249	16	9	solving	solve	VERB
cril-249	16	10	these	these	DET
cril-249	16	11	problems	problem	NOUN
cril-249	16	12	for	for	ADP
cril-249	16	13	arabic	arabic	ADJ
cril-249	16	14	and	and	CCONJ
cril-249	16	15	russian	russian	ADJ
cril-249	16	16	words	word	NOUN
cril-249	16	17	in	in	ADP
cril-249	16	18	our	our	PRON
cril-249	16	19	web	web	NOUN
cril-249	16	20	scraped	scrape	VERB
cril-249	16	21	corpus	corpu	NOUN
cril-249	16	22	.	.	PUNCT
cril-249	17	1	words	word	NOUN
cril-249	17	2	from	from	ADP
cril-249	17	3	both	both	PRON
cril-249	17	4	of	of	ADP
cril-249	17	5	these	these	DET
cril-249	17	6	languages	language	NOUN
cril-249	17	7	had	have	VERB
cril-249	17	8	1	1	NUM
cril-249	17	9	wrong	wrong	ADJ
cril-249	17	10	letter	letter	NOUN
cril-249	17	11	to	to	PART
cril-249	17	12	sound	sound	VERB
cril-249	17	13	output	output	NOUN
cril-249	17	14	of	of	ADP
cril-249	17	15	arabic	arabic	ADJ
cril-249	17	16	word	word	NOUN
cril-249	17	17	expanded	expand	VERB
cril-249	17	18	into	into	ADP
cril-249	17	19	arbabet	arbabet	ADJ
cril-249	17	20	symbols	symbol	NOUN
cril-249	17	21	using	use	VERB
cril-249	17	22	decision	decision	NOUN
cril-249	17	23	trees	tree	NOUN
cril-249	17	24	trained	train	VERB
cril-249	17	25	on	on	ADP
cril-249	17	26	cmu	cmu	PROPN
cril-249	17	27	’s	’s	PROPN
cril-249	17	28	english	english	PROPN
cril-249	17	29	lexicon	lexicon	PROPN
cril-249	17	30	.	.	PUNCT
cril-249	18	1	for	for	ADP
cril-249	18	2	details	detail	NOUN
cril-249	18	3	about	about	ADP
cril-249	18	4	arpabet	arpabet	ADJ
cril-249	18	5	symbols	symbol	NOUN
cril-249	18	6	,	,	PUNCT
cril-249	18	7	see	see	VERB
cril-249	18	8	(	(	PUNCT
cril-249	18	9	jurafsky	jurafsky	PROPN
cril-249	18	10	&	&	CCONJ
cril-249	18	11	martin	martin	PROPN
cril-249	18	12	2000	2000	NUM
cril-249	18	13	,	,	PUNCT
cril-249	18	14	p	p	NOUN
cril-249	18	15	94	94	NUM
cril-249	18	16	-	-	SYM
cril-249	18	17	95	95	NUM
cril-249	18	18	)	)	PUNCT
cril-249	18	19	.	.	PUNCT
cril-249	19	1	arpabet	arpabet	PROPN
cril-249	20	1	[	[	X
cril-249	20	2	hh	hh	PROPN
cril-249	20	3	ay	ay	PROPN
cril-249	20	4	b	b	PROPN
cril-249	20	5	ae	ae	PROPN
cril-249	20	6	kd	kd	PROPN
cril-249	20	7	]	]	X
cril-249	20	8	is	be	AUX
cril-249	20	9	roughly	roughly	ADV
cril-249	20	10	equivalent	equivalent	ADJ
cril-249	20	11	to	to	AUX
cril-249	20	12	ipa	ipa	PROPN
cril-249	20	13	[	[	X
cril-249	20	14	�	�	PROPN
cril-249	20	15	�	�	PROPN
cril-249	20	16	�	�	PROPN
cril-249	20	17	�	�	PROPN
cril-249	20	18	�	�	PROPN
cril-249	20	19	�	�	PROPN
cril-249	20	20	]	]	PUNCT
cril-249	20	21	.	.	PUNCT
cril-249	21	1	colorado	colorado	PROPN
cril-249	21	2	research	research	PROPN
cril-249	21	3	in	in	ADP
cril-249	21	4	linguistics	linguistic	NOUN
cril-249	21	5	.	.	PUNCT
cril-249	22	1	june	june	PROPN
cril-249	22	2	2004	2004	NUM
cril-249	22	3	.	.	PUNCT
cril-249	23	1	volume	volume	NOUN
cril-249	23	2	17	17	NUM
cril-249	23	3	,	,	PUNCT
cril-249	23	4	issue	issue	NOUN
cril-249	23	5	1	1	NUM
cril-249	23	6	.	.	PUNCT
cril-249	23	7	boulder	boulder	NOUN
cril-249	23	8	:	:	PUNCT
cril-249	23	9	university	university	NOUN
cril-249	23	10	of	of	ADP
cril-249	23	11	colorado	colorado	PROPN
cril-249	23	12	.	.	PUNCT
cril-249	24	1	©	©	PROPN
cril-249	24	2	2004	2004	NUM
cril-249	24	3	by	by	ADP
cril-249	24	4	lewis	lewis	PROPN
cril-249	24	5	,	,	PUNCT
cril-249	24	6	stephen	stephen	PROPN
cril-249	24	7	;	;	PUNCT
cril-249	24	8	mcgrath	mcgrath	PROPN
cril-249	24	9	,	,	PUNCT
cril-249	24	10	katie	katie	PROPN
cril-249	24	11	;	;	PUNCT
cril-249	24	12	reuppel	reuppel	PROPN
cril-249	24	13	,	,	PUNCT
cril-249	24	14	jeffrey	jeffrey	PROPN
cril-249	24	15	.	.	PROPN
cril-249	24	16	1	1	NUM
cril-249	24	17	lewis	lewis	PROPN
cril-249	24	18	et	et	PROPN
cril-249	24	19	al	al	PROPN
cril-249	24	20	.	.	PUNCT
cril-249	24	21	:	:	PUNCT
cril-249	25	1	language	language	NOUN
cril-249	25	2	identification	identification	NOUN
cril-249	25	3	and	and	CCONJ
cril-249	25	4	language	language	NOUN
cril-249	25	5	specific	specific	ADJ
cril-249	25	6	letter	letter	NOUN
cril-249	25	7	-	-	PUNCT
cril-249	25	8	to	to	ADP
cril-249	25	9	-	-	PUNCT
cril-249	25	10	sound	sound	NOUN
cril-249	25	11	rules	rule	NOUN
cril-249	25	12	published	publish	VERB
cril-249	25	13	by	by	ADP
cril-249	25	14	cu	cu	PROPN
cril-249	25	15	scholar	scholar	NOUN
cril-249	25	16	,	,	PUNCT
cril-249	25	17	2004	2004	NUM
cril-249	25	18	colorado	colorado	PROPN
cril-249	25	19	research	research	NOUN
cril-249	25	20	in	in	ADP
cril-249	25	21	linguistics	linguistic	NOUN
cril-249	25	22	,	,	PUNCT
cril-249	25	23	volume	volume	NOUN
cril-249	25	24	17	17	NUM
cril-249	25	25	(	(	PUNCT
cril-249	25	26	2004	2004	NUM
cril-249	25	27	)	)	PUNCT
cril-249	25	28	2	2	NUM
cril-249	25	29	already	already	ADV
cril-249	25	30	been	be	AUX
cril-249	25	31	identified	identify	VERB
cril-249	25	32	as	as	ADP
cril-249	25	33	both	both	CCONJ
cril-249	25	34	common	common	ADJ
cril-249	25	35	and	and	CCONJ
cril-249	25	36	problematic	problematic	ADJ
cril-249	25	37	to	to	ADP
cril-249	25	38	the	the	DET
cril-249	25	39	production	production	NOUN
cril-249	25	40	of	of	ADP
cril-249	25	41	training	training	NOUN
cril-249	25	42	material	material	NOUN
cril-249	25	43	.	.	PUNCT
cril-249	26	1	2	2	X
cril-249	26	2	.	.	X
cril-249	26	3	past	past	ADJ
cril-249	26	4	/	/	SYM
cril-249	26	5	related	related	ADJ
cril-249	26	6	work	work	NOUN
cril-249	26	7	off	off	ADP
cril-249	26	8	the	the	DET
cril-249	26	9	shelf	shelf	NOUN
cril-249	26	10	technology	technology	NOUN
cril-249	26	11	which	which	PRON
cril-249	26	12	can	can	AUX
cril-249	26	13	perform	perform	VERB
cril-249	26	14	language	language	NOUN
cril-249	26	15	identification	identification	NOUN
cril-249	26	16	on	on	ADP
cril-249	26	17	text	text	NOUN
cril-249	26	18	in	in	ADP
cril-249	26	19	general	general	ADJ
cril-249	26	20	is	be	AUX
cril-249	26	21	available	available	ADJ
cril-249	26	22	.	.	PUNCT
cril-249	27	1	these	these	DET
cril-249	27	2	same	same	ADJ
cril-249	27	3	tools	tool	NOUN
cril-249	27	4	can	can	AUX
cril-249	27	5	also	also	ADV
cril-249	27	6	be	be	AUX
cril-249	27	7	used	use	VERB
cril-249	27	8	to	to	PART
cril-249	27	9	produce	produce	VERB
cril-249	27	10	letter	letter	NOUN
cril-249	27	11	to	to	ADP
cril-249	27	12	speech	speech	NOUN
cril-249	27	13	pronunciation	pronunciation	NOUN
cril-249	27	14	output	output	NOUN
cril-249	27	15	for	for	ADP
cril-249	27	16	non	non	ADJ
cril-249	27	17	-	-	ADJ
cril-249	27	18	english	english	ADJ
cril-249	27	19	words	word	NOUN
cril-249	27	20	.	.	PUNCT
cril-249	28	1	however	however	ADV
cril-249	28	2	none	none	NOUN
cril-249	28	3	of	of	ADP
cril-249	28	4	these	these	DET
cril-249	28	5	systems	system	NOUN
cril-249	28	6	are	be	AUX
cril-249	28	7	appropriate	appropriate	ADJ
cril-249	28	8	for	for	ADP
cril-249	28	9	the	the	DET
cril-249	28	10	task	task	NOUN
cril-249	28	11	that	that	PRON
cril-249	28	12	we	we	PRON
cril-249	28	13	wish	wish	VERB
cril-249	28	14	to	to	PART
cril-249	28	15	accomplish	accomplish	VERB
cril-249	28	16	.	.	PUNCT
cril-249	29	1	most	most	ADJ
cril-249	29	2	existing	exist	VERB
cril-249	29	3	language	language	NOUN
cril-249	29	4	classification	classification	NOUN
cril-249	29	5	systems	system	NOUN
cril-249	29	6	use	use	VERB
cril-249	29	7	the	the	DET
cril-249	29	8	simple	simple	ADJ
cril-249	29	9	trick	trick	NOUN
cril-249	29	10	of	of	ADP
cril-249	29	11	determining	determine	VERB
cril-249	29	12	the	the	DET
cril-249	29	13	character	character	NOUN
cril-249	29	14	encoding	encoding	NOUN
cril-249	29	15	of	of	ADP
cril-249	29	16	a	a	DET
cril-249	29	17	document	document	NOUN
cril-249	29	18	to	to	PART
cril-249	29	19	perform	perform	VERB
cril-249	29	20	language	language	NOUN
cril-249	29	21	identification	identification	NOUN
cril-249	29	22	.	.	PUNCT
cril-249	30	1	a	a	DET
cril-249	30	2	few	few	ADJ
cril-249	30	3	systems	system	NOUN
cril-249	30	4	use	use	VERB
cril-249	30	5	n	n	CCONJ
cril-249	30	6	-	-	PUNCT
cril-249	30	7	gram	gram	NOUN
cril-249	30	8	statistics	statistic	NOUN
cril-249	30	9	at	at	ADP
cril-249	30	10	the	the	DET
cril-249	30	11	word	word	NOUN
cril-249	30	12	level	level	NOUN
cril-249	30	13	to	to	PART
cril-249	30	14	perform	perform	VERB
cril-249	30	15	this	this	DET
cril-249	30	16	task	task	NOUN
cril-249	30	17	.	.	PUNCT
cril-249	31	1	both	both	PRON
cril-249	31	2	of	of	ADP
cril-249	31	3	these	these	DET
cril-249	31	4	techniques	technique	NOUN
cril-249	31	5	can	can	AUX
cril-249	31	6	be	be	AUX
cril-249	31	7	used	use	VERB
cril-249	31	8	to	to	PART
cril-249	31	9	identify	identify	VERB
cril-249	31	10	non	non	ADJ
cril-249	31	11	-	-	ADJ
cril-249	31	12	english	english	ADJ
cril-249	31	13	words	word	NOUN
cril-249	31	14	successfully	successfully	ADV
cril-249	31	15	.	.	PUNCT
cril-249	32	1	however	however	ADV
cril-249	32	2	this	this	DET
cril-249	32	3	identification	identification	NOUN
cril-249	32	4	works	work	VERB
cril-249	32	5	within	within	ADP
cril-249	32	6	the	the	DET
cril-249	32	7	context	context	NOUN
cril-249	32	8	of	of	ADP
cril-249	32	9	the	the	DET
cril-249	32	10	entirety	entirety	NOUN
cril-249	32	11	of	of	ADP
cril-249	32	12	the	the	DET
cril-249	32	13	text	text	NOUN
cril-249	32	14	in	in	ADP
cril-249	32	15	question	question	NOUN
cril-249	32	16	being	be	AUX
cril-249	32	17	composed	compose	VERB
cril-249	32	18	in	in	ADP
cril-249	32	19	the	the	DET
cril-249	32	20	non	non	ADJ
cril-249	32	21	-	-	ADJ
cril-249	32	22	english	english	ADJ
cril-249	32	23	language	language	NOUN
cril-249	32	24	.	.	PUNCT
cril-249	33	1	this	this	DET
cril-249	33	2	type	type	NOUN
cril-249	33	3	of	of	ADP
cril-249	33	4	system	system	NOUN
cril-249	33	5	is	be	AUX
cril-249	33	6	not	not	PART
cril-249	33	7	optimized	optimize	VERB
cril-249	33	8	to	to	PART
cril-249	33	9	deal	deal	VERB
cril-249	33	10	with	with	ADP
cril-249	33	11	identifying	identify	VERB
cril-249	33	12	the	the	DET
cril-249	33	13	origin	origin	NOUN
cril-249	33	14	of	of	ADP
cril-249	33	15	non	non	ADJ
cril-249	33	16	-	-	ADJ
cril-249	33	17	english	english	ADJ
cril-249	33	18	words	word	NOUN
cril-249	33	19	within	within	ADP
cril-249	33	20	an	an	DET
cril-249	33	21	otherwise	otherwise	ADV
cril-249	33	22	english	english	ADJ
cril-249	33	23	language	language	NOUN
cril-249	33	24	text	text	NOUN
cril-249	33	25	.	.	PUNCT
cril-249	34	1	the	the	DET
cril-249	34	2	same	same	ADJ
cril-249	34	3	is	be	AUX
cril-249	34	4	often	often	ADV
cril-249	34	5	true	true	ADJ
cril-249	34	6	of	of	ADP
cril-249	34	7	letter	letter	NOUN
cril-249	34	8	-	-	PUNCT
cril-249	34	9	to	to	ADP
cril-249	34	10	-	-	PUNCT
cril-249	34	11	sound	sound	NOUN
cril-249	34	12	systems	system	NOUN
cril-249	34	13	.	.	PUNCT
cril-249	35	1	they	they	PRON
cril-249	35	2	are	be	AUX
cril-249	35	3	built	build	VERB
cril-249	35	4	to	to	PART
cril-249	35	5	pronounce	pronounce	VERB
cril-249	35	6	nonenglish	nonenglish	ADJ
cril-249	35	7	words	word	NOUN
cril-249	35	8	correctly	correctly	ADV
cril-249	35	9	within	within	ADP
cril-249	35	10	the	the	DET
cril-249	35	11	context	context	NOUN
cril-249	35	12	of	of	ADP
cril-249	35	13	the	the	DET
cril-249	35	14	language	language	NOUN
cril-249	35	15	of	of	ADP
cril-249	35	16	origin	origin	NOUN
cril-249	35	17	.	.	PUNCT
cril-249	36	1	however	however	ADV
cril-249	36	2	the	the	DET
cril-249	36	3	pronunciations	pronunciation	NOUN
cril-249	36	4	are	be	AUX
cril-249	36	5	representative	representative	ADJ
cril-249	36	6	of	of	ADP
cril-249	36	7	native	native	ADJ
cril-249	36	8	speaker	speaker	NOUN
cril-249	36	9	pronunciation	pronunciation	NOUN
cril-249	36	10	and	and	CCONJ
cril-249	36	11	inappropriate	inappropriate	ADJ
cril-249	36	12	to	to	ADP
cril-249	36	13	producing	produce	VERB
cril-249	36	14	the	the	DET
cril-249	36	15	letter	letter	NOUN
cril-249	36	16	to	to	PART
cril-249	36	17	sound	sound	VERB
cril-249	36	18	output	output	NOUN
cril-249	36	19	indicative	indicative	ADJ
cril-249	36	20	of	of	ADP
cril-249	36	21	a	a	DET
cril-249	36	22	monolingual	monolingual	ADJ
cril-249	36	23	english	english	ADJ
cril-249	36	24	speaker	speaker	NOUN
cril-249	36	25	attempting	attempt	VERB
cril-249	36	26	to	to	PART
cril-249	36	27	pronounce	pronounce	VERB
cril-249	36	28	an	an	DET
cril-249	36	29	unknown	unknown	ADJ
cril-249	36	30	word	word	NOUN
cril-249	36	31	of	of	ADP
cril-249	36	32	foreign	foreign	ADJ
cril-249	36	33	origin	origin	NOUN
cril-249	36	34	.	.	PUNCT
cril-249	37	1	3	3	X
cril-249	37	2	.	.	X
cril-249	37	3	language	language	NOUN
cril-249	37	4	classification	classification	NOUN
cril-249	37	5	3.1	3.1	NUM
cril-249	37	6	.	.	PUNCT
cril-249	38	1	training	training	NOUN
cril-249	38	2	data	datum	NOUN
cril-249	38	3	since	since	SCONJ
cril-249	38	4	arabic	arabic	ADJ
cril-249	38	5	and	and	CCONJ
cril-249	38	6	russian	russian	ADJ
cril-249	38	7	proper	proper	ADJ
cril-249	38	8	nouns	noun	NOUN
cril-249	38	9	had	have	AUX
cril-249	38	10	previously	previously	ADV
cril-249	38	11	been	be	AUX
cril-249	38	12	identified	identify	VERB
cril-249	38	13	as	as	ADP
cril-249	38	14	the	the	DET
cril-249	38	15	primary	primary	ADJ
cril-249	38	16	cause	cause	NOUN
cril-249	38	17	for	for	ADP
cril-249	38	18	pronunciation	pronunciation	NOUN
cril-249	38	19	errors	error	NOUN
cril-249	38	20	our	our	PRON
cril-249	38	21	first	first	ADJ
cril-249	38	22	task	task	NOUN
cril-249	38	23	was	be	AUX
cril-249	38	24	to	to	PART
cril-249	38	25	acquire	acquire	VERB
cril-249	38	26	training	training	NOUN
cril-249	38	27	data	datum	NOUN
cril-249	38	28	by	by	ADP
cril-249	38	29	which	which	PRON
cril-249	38	30	to	to	PART
cril-249	38	31	identify	identify	VERB
cril-249	38	32	these	these	DET
cril-249	38	33	words	word	NOUN
cril-249	38	34	.	.	PUNCT
cril-249	39	1	using	use	VERB
cril-249	39	2	the	the	DET
cril-249	39	3	world	world	NOUN
cril-249	39	4	wide	wide	ADJ
cril-249	39	5	web	web	NOUN
cril-249	39	6	as	as	ADP
cril-249	39	7	our	our	PRON
cril-249	39	8	resource	resource	NOUN
cril-249	39	9	we	we	PRON
cril-249	39	10	constructed	construct	VERB
cril-249	39	11	a	a	DET
cril-249	39	12	database	database	NOUN
cril-249	39	13	of	of	ADP
cril-249	39	14	transliterated	transliterated	ADJ
cril-249	39	15	proper	proper	ADJ
cril-249	39	16	nouns	noun	NOUN
cril-249	39	17	in	in	ADP
cril-249	39	18	arabic	arabic	ADJ
cril-249	39	19	and	and	CCONJ
cril-249	39	20	russian	russian	ADJ
cril-249	39	21	names	name	NOUN
cril-249	39	22	.	.	PUNCT
cril-249	40	1	the	the	DET
cril-249	40	2	arabic	arabic	ADJ
cril-249	40	3	names	name	NOUN
cril-249	40	4	were	be	AUX
cril-249	40	5	obtained	obtain	VERB
cril-249	40	6	on	on	ADP
cril-249	40	7	the	the	DET
cril-249	40	8	web	web	NOUN
cril-249	40	9	in	in	ADP
cril-249	40	10	numerous	numerous	ADJ
cril-249	40	11	locations	location	NOUN
cril-249	40	12	to	to	PART
cril-249	40	13	create	create	VERB
cril-249	40	14	a	a	DET
cril-249	40	15	list	list	NOUN
cril-249	40	16	of	of	ADP
cril-249	40	17	3143	3143	NUM
cril-249	40	18	unique	unique	ADJ
cril-249	40	19	arabic	arabic	ADJ
cril-249	40	20	names	name	NOUN
cril-249	40	21	.	.	PUNCT
cril-249	41	1	the	the	DET
cril-249	41	2	russian	russian	ADJ
cril-249	41	3	names	name	NOUN
cril-249	41	4	were	be	AUX
cril-249	41	5	all	all	PRON
cril-249	41	6	obtained	obtain	VERB
cril-249	41	7	from	from	ADP
cril-249	41	8	a	a	DET
cril-249	41	9	single	single	ADJ
cril-249	41	10	web	web	NOUN
cril-249	41	11	site	site	NOUN
cril-249	41	12	that	that	PRON
cril-249	41	13	provided	provide	VERB
cril-249	41	14	20577	20577	NUM
cril-249	41	15	unique	unique	ADJ
cril-249	41	16	russian	russian	ADJ
cril-249	41	17	names	name	NOUN
cril-249	41	18	(	(	PUNCT
cril-249	41	19	goldschmidt	goldschmidt	NOUN
cril-249	41	20	1996	1996	NUM
cril-249	41	21	)	)	PUNCT
cril-249	41	22	.	.	PUNCT
cril-249	42	1	in	in	ADP
cril-249	42	2	order	order	NOUN
cril-249	42	3	to	to	PART
cril-249	42	4	create	create	VERB
cril-249	42	5	a	a	DET
cril-249	42	6	standard	standard	NOUN
cril-249	42	7	of	of	ADP
cril-249	42	8	comparison	comparison	NOUN
cril-249	42	9	for	for	ADP
cril-249	42	10	english	english	ADJ
cril-249	42	11	words	word	NOUN
cril-249	42	12	,	,	PUNCT
cril-249	42	13	we	we	PRON
cril-249	42	14	also	also	ADV
cril-249	42	15	collected	collect	VERB
cril-249	42	16	a	a	DET
cril-249	42	17	list	list	NOUN
cril-249	42	18	of	of	ADP
cril-249	42	19	the	the	DET
cril-249	42	20	10,000	10,000	NUM
cril-249	42	21	most	most	ADV
cril-249	42	22	common	common	ADJ
cril-249	42	23	words	word	NOUN
cril-249	42	24	longer	long	ADV
cril-249	42	25	than	than	ADP
cril-249	42	26	4	4	NUM
cril-249	42	27	letters	letter	NOUN
cril-249	42	28	from	from	ADP
cril-249	42	29	the	the	DET
cril-249	42	30	brown	brown	ADJ
cril-249	42	31	corpus	corpus	NOUN
cril-249	42	32	.	.	PUNCT
cril-249	43	1	we	we	PRON
cril-249	43	2	collected	collect	VERB
cril-249	43	3	english	english	ADJ
cril-249	43	4	words	word	NOUN
cril-249	43	5	from	from	ADP
cril-249	43	6	the	the	DET
cril-249	43	7	brown	brown	ADJ
cril-249	43	8	corpus	corpus	NOUN
cril-249	43	9	with	with	ADP
cril-249	43	10	the	the	DET
cril-249	43	11	belief	belief	NOUN
cril-249	43	12	that	that	SCONJ
cril-249	43	13	such	such	DET
cril-249	43	14	an	an	DET
cril-249	43	15	assemblage	assemblage	NOUN
cril-249	43	16	would	would	AUX
cril-249	43	17	better	well	ADV
cril-249	43	18	represent	represent	VERB
cril-249	43	19	the	the	DET
cril-249	43	20	body	body	NOUN
cril-249	43	21	of	of	ADP
cril-249	43	22	unknown	unknown	ADJ
cril-249	43	23	english	english	ADJ
cril-249	43	24	words	word	NOUN
cril-249	43	25	.	.	PUNCT
cril-249	44	1	while	while	SCONJ
cril-249	44	2	new	new	ADJ
cril-249	44	3	foreign	foreign	ADJ
cril-249	44	4	words	word	NOUN
cril-249	44	5	are	be	AUX
cril-249	44	6	generally	generally	ADV
cril-249	44	7	proper	proper	ADJ
cril-249	44	8	nouns	noun	NOUN
cril-249	44	9	,	,	PUNCT
cril-249	44	10	"	"	PUNCT
cril-249	44	11	unknown	unknown	ADJ
cril-249	44	12	"	"	PUNCT
cril-249	44	13	english	english	ADJ
cril-249	44	14	words	word	NOUN
cril-249	44	15	tend	tend	VERB
cril-249	44	16	to	to	PART
cril-249	44	17	be	be	AUX
cril-249	44	18	morphological	morphological	ADJ
cril-249	44	19	variants	variant	NOUN
cril-249	44	20	of	of	ADP
cril-249	44	21	words	word	NOUN
cril-249	44	22	in	in	ADP
cril-249	44	23	the	the	DET
cril-249	44	24	lexicon	lexicon	NOUN
cril-249	44	25	.	.	PUNCT
cril-249	45	1	in	in	ADP
cril-249	45	2	addition	addition	NOUN
cril-249	45	3	,	,	PUNCT
cril-249	45	4	english	english	PROPN
cril-249	45	5	first	first	ADJ
cril-249	45	6	names	name	NOUN
cril-249	45	7	are	be	AUX
cril-249	45	8	not	not	PART
cril-249	45	9	at	at	ADV
cril-249	45	10	all	all	ADV
cril-249	45	11	representative	representative	NOUN
cril-249	45	12	of	of	ADP
cril-249	45	13	the	the	DET
cril-249	45	14	greater	great	ADJ
cril-249	45	15	language	language	NOUN
cril-249	45	16	,	,	PUNCT
cril-249	45	17	as	as	SCONJ
cril-249	45	18	evidenced	evidence	VERB
cril-249	45	19	by	by	ADP
cril-249	45	20	the	the	DET
cril-249	45	21	5	5	NUM
cril-249	45	22	most	most	ADV
cril-249	45	23	common	common	ADJ
cril-249	45	24	4	4	NUM
cril-249	45	25	-	-	PUNCT
cril-249	45	26	grams	gram	NOUN
cril-249	45	27	from	from	ADP
cril-249	45	28	a	a	DET
cril-249	45	29	list	list	NOUN
cril-249	45	30	of	of	ADP
cril-249	45	31	english	english	ADJ
cril-249	45	32	names	name	NOUN
cril-249	45	33	and	and	CCONJ
cril-249	45	34	the	the	DET
cril-249	45	35	brown	brown	ADJ
cril-249	45	36	corpus	corpus	NOUN
cril-249	45	37	(	(	PUNCT
cril-249	45	38	table	table	NOUN
cril-249	45	39	1	1	NUM
cril-249	45	40	)	)	PUNCT
cril-249	45	41	.	.	PUNCT
cril-249	46	1	2	2	NUM
cril-249	46	2	colorado	colorado	NOUN
cril-249	46	3	research	research	NOUN
cril-249	46	4	in	in	ADP
cril-249	46	5	linguistics	linguistic	NOUN
cril-249	46	6	,	,	PUNCT
cril-249	46	7	vol	vol	NOUN
cril-249	46	8	.	.	PROPN
cril-249	46	9	17	17	NUM
cril-249	47	1	[	[	X
cril-249	47	2	2004	2004	NUM
cril-249	47	3	]	]	PUNCT
cril-249	47	4	https://scholar.colorado.edu/cril/vol17/iss1/6	https://scholar.colorado.edu/cril/vol17/iss1/6	NOUN
cril-249	47	5	doi	doi	NOUN
cril-249	47	6	:	:	PUNCT
cril-249	47	7	https://doi.org/10.25810/60mf-fn94	https://doi.org/10.25810/60mf-fn94	PROPN
cril-249	47	8	http://www.sca.org/heraldry/paul/	http://www.sca.org/heraldry/paul/	PROPN
cril-249	47	9	language	language	NOUN
cril-249	47	10	identification	identification	NOUN
cril-249	47	11	and	and	CCONJ
cril-249	47	12	language	language	NOUN
cril-249	47	13	specific	specific	ADJ
cril-249	47	14	letter	letter	NOUN
cril-249	47	15	-	-	PUNCT
cril-249	47	16	to	to	ADP
cril-249	47	17	-	-	PUNCT
cril-249	47	18	sound	sound	NOUN
cril-249	47	19	rules	rule	NOUN
cril-249	47	20	3	3	NUM
cril-249	47	21	names	name	NOUN
cril-249	47	22	brown	brown	ADJ
cril-249	47	23	corpus	corpus	NOUN
cril-249	47	24	<	<	X
cril-249	47	25	s	s	PROPN
cril-249	47	26	>	>	X
cril-249	47	27	mar	mar	PROPN
cril-249	47	28	ing</s	ing</s	PROPN
cril-249	47	29	>	>	X
cril-249	47	30	ana</s	ana</s	PROPN
cril-249	47	31	>	>	X
cril-249	47	32	tion	tion	PROPN
cril-249	47	33	nna</s	nna</s	PROPN
cril-249	47	34	>	>	X
cril-249	47	35	ion</s	ion</s	X
cril-249	47	36	>	>	X
cril-249	47	37	ina</s	ina</s	PROPN
cril-249	47	38	>	>	X
cril-249	47	39	atio	atio	PROPN
cril-249	47	40	anna	anna	PROPN
cril-249	47	41	ted</s	ted</s	PROPN
cril-249	47	42	>	>	PROPN
cril-249	47	43	table	table	NOUN
cril-249	47	44	1	1	NUM
cril-249	47	45	.	.	PUNCT
cril-249	48	1	the	the	DET
cril-249	48	2	5	5	NUM
cril-249	48	3	most	most	ADV
cril-249	48	4	common	common	ADJ
cril-249	48	5	4	4	NUM
cril-249	48	6	-	-	PUNCT
cril-249	48	7	grams	gram	NOUN
cril-249	48	8	from	from	ADP
cril-249	48	9	a	a	DET
cril-249	48	10	list	list	NOUN
cril-249	48	11	of	of	ADP
cril-249	48	12	english	english	ADJ
cril-249	48	13	names	name	NOUN
cril-249	48	14	and	and	CCONJ
cril-249	48	15	from	from	ADP
cril-249	48	16	the	the	DET
cril-249	48	17	brown	brown	ADJ
cril-249	48	18	corpus	corpus	PROPN
cril-249	48	19	3.2	3.2	NUM
cril-249	48	20	.	.	PUNCT
cril-249	49	1	classification	classification	NOUN
cril-249	49	2	algorithm	algorithm	NOUN
cril-249	49	3	we	we	PRON
cril-249	49	4	implemented	implement	VERB
cril-249	49	5	an	an	DET
cril-249	49	6	n	n	CCONJ
cril-249	49	7	-	-	PUNCT
cril-249	49	8	gram	gram	NOUN
cril-249	49	9	classifier	classifier	NOUN
cril-249	49	10	to	to	PART
cril-249	49	11	handle	handle	VERB
cril-249	49	12	language	language	NOUN
cril-249	49	13	type	type	NOUN
cril-249	49	14	identification	identification	NOUN
cril-249	49	15	.	.	PUNCT
cril-249	50	1	each	each	DET
cril-249	50	2	training	training	NOUN
cril-249	50	3	word	word	NOUN
cril-249	50	4	was	be	AUX
cril-249	50	5	segmented	segment	VERB
cril-249	50	6	into	into	ADP
cril-249	50	7	individual	individual	ADJ
cril-249	50	8	letters	letter	NOUN
cril-249	50	9	.	.	PUNCT
cril-249	51	1	individual	individual	ADJ
cril-249	51	2	4	4	NUM
cril-249	51	3	-	-	PUNCT
cril-249	51	4	grams	gram	NOUN
cril-249	51	5	were	be	AUX
cril-249	51	6	constructed	construct	VERB
cril-249	51	7	using	use	VERB
cril-249	51	8	each	each	DET
cril-249	51	9	four	four	NUM
cril-249	51	10	-	-	PUNCT
cril-249	51	11	letter	letter	NOUN
cril-249	51	12	set	set	NOUN
cril-249	51	13	.	.	PUNCT
cril-249	52	1	in	in	ADP
cril-249	52	2	addition	addition	NOUN
cril-249	52	3	much	much	ADV
cril-249	52	4	like	like	ADP
cril-249	52	5	sentence	sentence	NOUN
cril-249	52	6	boundaries	boundary	NOUN
cril-249	52	7	are	be	AUX
cril-249	52	8	marked	mark	VERB
cril-249	52	9	,	,	PUNCT
cril-249	52	10	letters	letter	NOUN
cril-249	52	11	which	which	PRON
cril-249	52	12	begin	begin	VERB
cril-249	52	13	and	and	CCONJ
cril-249	52	14	end	end	VERB
cril-249	52	15	words	word	NOUN
cril-249	52	16	were	be	AUX
cril-249	52	17	marked	mark	VERB
cril-249	52	18	with	with	ADP
cril-249	52	19	<	<	X
cril-249	52	20	s	s	X
cril-249	52	21	>	>	X
cril-249	52	22	and	and	CCONJ
cril-249	52	23	<	<	X
cril-249	52	24	\s	\s	PROPN
cril-249	52	25	>	>	X
cril-249	52	26	.	.	PUNCT
cril-249	53	1	each	each	DET
cril-249	53	2	individual	individual	ADJ
cril-249	53	3	4	4	NUM
cril-249	53	4	-	-	PUNCT
cril-249	53	5	gram	gram	NOUN
cril-249	53	6	was	be	AUX
cril-249	53	7	then	then	ADV
cril-249	53	8	assigned	assign	VERB
cril-249	53	9	a	a	DET
cril-249	53	10	specific	specific	ADJ
cril-249	53	11	probability	probability	NOUN
cril-249	53	12	based	base	VERB
cril-249	53	13	on	on	ADP
cril-249	53	14	frequency	frequency	NOUN
cril-249	53	15	.	.	PUNCT
cril-249	54	1	the	the	DET
cril-249	54	2	word	word	NOUN
cril-249	54	3	to	to	PART
cril-249	54	4	be	be	AUX
cril-249	54	5	classified	classify	VERB
cril-249	54	6	was	be	AUX
cril-249	54	7	also	also	ADV
cril-249	54	8	segmented	segment	VERB
cril-249	54	9	into	into	ADP
cril-249	54	10	4	4	NUM
cril-249	54	11	-	-	PUNCT
cril-249	54	12	grams	gram	NOUN
cril-249	54	13	and	and	CCONJ
cril-249	54	14	then	then	ADV
cril-249	54	15	labeled	label	VERB
cril-249	54	16	for	for	ADP
cril-249	54	17	language	language	NOUN
cril-249	54	18	using	use	VERB
cril-249	54	19	the	the	DET
cril-249	54	20	following	follow	VERB
cril-249	54	21	equation	equation	NOUN
cril-249	54	22	(	(	PUNCT
cril-249	54	23	equation	equation	NOUN
cril-249	54	24	1	1	NUM
cril-249	54	25	)	)	PUNCT
cril-249	54	26	.	.	PUNCT
cril-249	55	1	equation	equation	NOUN
cril-249	55	2	1	1	NUM
cril-249	55	3	:	:	PUNCT
cril-249	55	4	c	c	NOUN
cril-249	55	5	=	=	PUNCT
cril-249	55	6	argmax	argmax	PROPN
cril-249	55	7	c	c	NOUN
cril-249	55	8	p(c|x1,	p(c|x1,	X
cril-249	55	9	…	…	PUNCT
cril-249	55	10	,xn	,xn	NUM
cril-249	55	11	)	)	PUNCT
cril-249	55	12	=	=	PUNCT
cril-249	55	13	argmax	argmax	PRON
cril-249	55	14	c	c	NOUN
cril-249	55	15	p(c	p(c	PROPN
cril-249	55	16	)	)	PUNCT
cril-249	55	17	π	π	PROPN
cril-249	55	18	p(xi|c	p(xi|c	PROPN
cril-249	55	19	)	)	PUNCT
cril-249	55	20	i=1	i=1	PROPN
cril-249	56	1	c	c	NOUN
cril-249	56	2	–	–	PUNCT
cril-249	56	3	final	final	ADJ
cril-249	56	4	language	language	NOUN
cril-249	56	5	classification	classification	NOUN
cril-249	56	6	c	c	NOUN
cril-249	56	7	–	–	PUNCT
cril-249	56	8	individual	individual	ADJ
cril-249	56	9	language	language	NOUN
cril-249	56	10	classification	classification	NOUN
cril-249	56	11	x	x	SYM
cril-249	56	12	–	–	PUNCT
cril-249	56	13	4	4	NUM
cril-249	56	14	-	-	PUNCT
cril-249	56	15	gram	gram	NOUN
cril-249	56	16	n	n	NOUN
cril-249	56	17	–	–	PUNCT
cril-249	56	18	number	number	NOUN
cril-249	56	19	of	of	ADP
cril-249	56	20	4	4	NUM
cril-249	56	21	-	-	PUNCT
cril-249	56	22	grams	gram	NOUN
cril-249	56	23	in	in	ADP
cril-249	56	24	the	the	DET
cril-249	56	25	word	word	NOUN
cril-249	56	26	being	be	AUX
cril-249	56	27	classified	classify	VERB
cril-249	56	28	prior	prior	ADJ
cril-249	56	29	probabilities	probability	NOUN
cril-249	56	30	for	for	ADP
cril-249	56	31	each	each	DET
cril-249	56	32	language	language	NOUN
cril-249	56	33	were	be	AUX
cril-249	56	34	generated	generate	VERB
cril-249	56	35	in	in	ADP
cril-249	56	36	proportion	proportion	NOUN
cril-249	56	37	to	to	ADP
cril-249	56	38	the	the	DET
cril-249	56	39	content	content	NOUN
cril-249	56	40	of	of	ADP
cril-249	56	41	the	the	DET
cril-249	56	42	cnn	cnn	PROPN
cril-249	56	43	corpus	corpus	PROPN
cril-249	56	44	.	.	PUNCT
cril-249	57	1	this	this	DET
cril-249	57	2	set	set	NOUN
cril-249	57	3	of	of	ADP
cril-249	57	4	news	news	NOUN
cril-249	57	5	transcriptions	transcription	NOUN
cril-249	57	6	was	be	AUX
cril-249	57	7	then	then	ADV
cril-249	57	8	compared	compare	VERB
cril-249	57	9	against	against	ADP
cril-249	57	10	the	the	DET
cril-249	57	11	cmu	cmu	PROPN
cril-249	57	12	lexicon	lexicon	PROPN
cril-249	57	13	.	.	PUNCT
cril-249	58	1	this	this	DET
cril-249	58	2	comparison	comparison	NOUN
cril-249	58	3	returned	return	VERB
cril-249	58	4	a	a	DET
cril-249	58	5	list	list	NOUN
cril-249	58	6	of	of	ADP
cril-249	58	7	1001	1001	NUM
cril-249	58	8	"	"	PUNCT
cril-249	58	9	unknown	unknown	ADJ
cril-249	58	10	"	"	PUNCT
cril-249	58	11	words	word	NOUN
cril-249	58	12	.	.	PUNCT
cril-249	59	1	each	each	DET
cril-249	59	2	unknown	unknown	ADJ
cril-249	59	3	word	word	NOUN
cril-249	59	4	was	be	AUX
cril-249	59	5	labeled	label	VERB
cril-249	59	6	by	by	ADP
cril-249	59	7	a	a	DET
cril-249	59	8	team	team	NOUN
cril-249	59	9	of	of	ADP
cril-249	59	10	linguistics	linguistic	NOUN
cril-249	59	11	graduate	graduate	NOUN
cril-249	59	12	students	student	NOUN
cril-249	59	13	as	as	ADP
cril-249	59	14	being	be	AUX
cril-249	59	15	arabic	arabic	ADJ
cril-249	59	16	,	,	PUNCT
cril-249	59	17	russian	russian	ADJ
cril-249	59	18	,	,	PUNCT
cril-249	59	19	or	or	CCONJ
cril-249	59	20	other	other	ADJ
cril-249	59	21	.	.	PUNCT
cril-249	60	1	each	each	DET
cril-249	60	2	word	word	NOUN
cril-249	60	3	was	be	AUX
cril-249	60	4	then	then	ADV
cril-249	60	5	classified	classify	VERB
cril-249	60	6	according	accord	VERB
cril-249	60	7	to	to	ADP
cril-249	60	8	the	the	DET
cril-249	60	9	majority	majority	NOUN
cril-249	60	10	label	label	NOUN
cril-249	60	11	given	give	VERB
cril-249	60	12	it	it	PRON
cril-249	60	13	and	and	CCONJ
cril-249	60	14	the	the	DET
cril-249	60	15	percentages	percentage	NOUN
cril-249	60	16	of	of	ADP
cril-249	60	17	each	each	DET
cril-249	60	18	language	language	NOUN
cril-249	60	19	classification	classification	NOUN
cril-249	60	20	determined	determine	VERB
cril-249	60	21	the	the	DET
cril-249	60	22	priors	prior	NOUN
cril-249	60	23	.	.	PUNCT
cril-249	61	1	the	the	DET
cril-249	61	2	priors	prior	NOUN
cril-249	61	3	were	be	AUX
cril-249	61	4	set	set	VERB
cril-249	61	5	as	as	ADP
cril-249	61	6	in	in	ADP
cril-249	61	7	table	table	NOUN
cril-249	61	8	2	2	NUM
cril-249	61	9	:	:	SYM
cril-249	61	10	3	3	NUM
cril-249	61	11	lewis	lewis	PROPN
cril-249	61	12	et	et	PROPN
cril-249	61	13	al	al	PROPN
cril-249	61	14	.	.	PUNCT
cril-249	61	15	:	:	PUNCT
cril-249	62	1	language	language	NOUN
cril-249	62	2	identification	identification	NOUN
cril-249	62	3	and	and	CCONJ
cril-249	62	4	language	language	NOUN
cril-249	62	5	specific	specific	ADJ
cril-249	62	6	letter	letter	NOUN
cril-249	62	7	-	-	PUNCT
cril-249	62	8	to	to	ADP
cril-249	62	9	-	-	PUNCT
cril-249	62	10	sound	sound	NOUN
cril-249	62	11	rules	rule	NOUN
cril-249	62	12	published	publish	VERB
cril-249	62	13	by	by	ADP
cril-249	62	14	cu	cu	PROPN
cril-249	62	15	scholar	scholar	NOUN
cril-249	62	16	,	,	PUNCT
cril-249	62	17	2004	2004	NUM
cril-249	62	18	colorado	colorado	PROPN
cril-249	62	19	research	research	NOUN
cril-249	62	20	in	in	ADP
cril-249	62	21	linguistics	linguistic	NOUN
cril-249	62	22	,	,	PUNCT
cril-249	62	23	volume	volume	NOUN
cril-249	62	24	17	17	NUM
cril-249	62	25	(	(	PUNCT
cril-249	62	26	2004	2004	NUM
cril-249	62	27	)	)	PUNCT
cril-249	62	28	4	4	NUM
cril-249	62	29	english	english	NOUN
cril-249	62	30	:	:	PUNCT
cril-249	62	31	0.805	0.805	NUM
cril-249	62	32	arabic	arabic	ADJ
cril-249	62	33	:	:	PUNCT
cril-249	62	34	0.156	0.156	NUM
cril-249	62	35	russian	russian	NOUN
cril-249	62	36	:	:	PUNCT
cril-249	62	37	0.039	0.039	NUM
cril-249	62	38	table	table	NOUN
cril-249	62	39	2	2	NUM
cril-249	62	40	.	.	PUNCT
cril-249	62	41	prior	prior	ADJ
cril-249	62	42	probabilities	probability	NOUN
cril-249	62	43	for	for	ADP
cril-249	62	44	labeling	labeling	NOUN
cril-249	62	45	words	word	NOUN
cril-249	62	46	english	english	PROPN
cril-249	62	47	,	,	PUNCT
cril-249	62	48	arabic	arabic	ADJ
cril-249	62	49	,	,	PUNCT
cril-249	62	50	or	or	CCONJ
cril-249	62	51	russian	russian	ADJ
cril-249	62	52	as	as	ADP
cril-249	62	53	with	with	ADP
cril-249	62	54	any	any	DET
cril-249	62	55	n	n	CCONJ
cril-249	62	56	-	-	PUNCT
cril-249	62	57	gram	gram	NOUN
cril-249	62	58	classifier	classifier	NOUN
cril-249	62	59	,	,	PUNCT
cril-249	62	60	we	we	PRON
cril-249	62	61	needed	need	VERB
cril-249	62	62	a	a	DET
cril-249	62	63	way	way	NOUN
cril-249	62	64	to	to	PART
cril-249	62	65	calculate	calculate	VERB
cril-249	62	66	probabilities	probability	NOUN
cril-249	62	67	for	for	ADP
cril-249	62	68	new	new	ADJ
cril-249	62	69	4grams	4gram	NOUN
cril-249	62	70	that	that	PRON
cril-249	62	71	never	never	ADV
cril-249	62	72	appeared	appear	VERB
cril-249	62	73	in	in	ADP
cril-249	62	74	the	the	DET
cril-249	62	75	training	training	NOUN
cril-249	62	76	data	datum	NOUN
cril-249	62	77	.	.	PUNCT
cril-249	63	1	to	to	PART
cril-249	63	2	account	account	VERB
cril-249	63	3	for	for	ADP
cril-249	63	4	these	these	DET
cril-249	63	5	unseen	unseen	ADJ
cril-249	63	6	4	4	NUM
cril-249	63	7	-	-	PUNCT
cril-249	63	8	grams	gram	NOUN
cril-249	63	9	,	,	PUNCT
cril-249	63	10	we	we	PRON
cril-249	63	11	used	use	VERB
cril-249	63	12	a	a	DET
cril-249	63	13	modified	modified	ADJ
cril-249	63	14	add	add	VERB
cril-249	63	15	-	-	PUNCT
cril-249	63	16	one	one	NUM
cril-249	63	17	smoothing	smoothing	NOUN
cril-249	63	18	method	method	NOUN
cril-249	63	19	using	use	VERB
cril-249	63	20	this	this	DET
cril-249	63	21	method	method	NOUN
cril-249	63	22	,	,	PUNCT
cril-249	63	23	the	the	DET
cril-249	63	24	probability	probability	NOUN
cril-249	63	25	for	for	ADP
cril-249	63	26	each	each	DET
cril-249	63	27	unseen	unseen	ADJ
cril-249	63	28	4	4	NUM
cril-249	63	29	-	-	PUNCT
cril-249	63	30	gram	gram	NOUN
cril-249	63	31	was	be	AUX
cril-249	63	32	assigned	assign	VERB
cril-249	63	33	to	to	PART
cril-249	63	34	be	be	AUX
cril-249	63	35	the	the	DET
cril-249	63	36	same	same	ADJ
cril-249	63	37	as	as	ADP
cril-249	63	38	that	that	PRON
cril-249	63	39	of	of	ADP
cril-249	63	40	the	the	DET
cril-249	63	41	4	4	NUM
cril-249	63	42	-	-	PUNCT
cril-249	63	43	grams	gram	NOUN
cril-249	63	44	with	with	ADP
cril-249	63	45	the	the	DET
cril-249	63	46	lowest	low	ADJ
cril-249	63	47	overall	overall	ADJ
cril-249	63	48	probability	probability	NOUN
cril-249	63	49	.	.	PUNCT
cril-249	64	1	in	in	ADP
cril-249	64	2	the	the	DET
cril-249	64	3	same	same	ADJ
cril-249	64	4	1001	1001	NUM
cril-249	64	5	unknown	unknown	ADJ
cril-249	64	6	words	word	NOUN
cril-249	64	7	mentioned	mention	VERB
cril-249	64	8	above	above	ADV
cril-249	64	9	,	,	PUNCT
cril-249	64	10	unseen	unseen	ADJ
cril-249	64	11	4	4	NUM
cril-249	64	12	-	-	PUNCT
cril-249	64	13	grams	gram	NOUN
cril-249	64	14	accounted	account	VERB
cril-249	64	15	for	for	ADP
cril-249	64	16	only	only	ADV
cril-249	64	17	0.126	0.126	NUM
cril-249	64	18	%	%	NOUN
cril-249	64	19	of	of	ADP
cril-249	64	20	all	all	DET
cril-249	64	21	4	4	NUM
cril-249	64	22	-	-	PUNCT
cril-249	64	23	grams	gram	NOUN
cril-249	64	24	.	.	PUNCT
cril-249	65	1	as	as	SCONJ
cril-249	65	2	smoothing	smooth	VERB
cril-249	65	3	is	be	AUX
cril-249	65	4	invoked	invoke	VERB
cril-249	65	5	so	so	ADV
cril-249	65	6	infrequently	infrequently	ADV
cril-249	65	7	,	,	PUNCT
cril-249	65	8	more	more	ADV
cril-249	65	9	sophisticated	sophisticated	ADJ
cril-249	65	10	forms	form	NOUN
cril-249	65	11	of	of	ADP
cril-249	65	12	smoothing	smoothing	NOUN
cril-249	65	13	or	or	CCONJ
cril-249	65	14	back	back	NOUN
cril-249	65	15	-	-	PUNCT
cril-249	65	16	off	off	NOUN
cril-249	65	17	do	do	AUX
cril-249	65	18	not	not	PART
cril-249	65	19	seem	seem	VERB
cril-249	65	20	to	to	PART
cril-249	65	21	be	be	AUX
cril-249	65	22	fertile	fertile	ADJ
cril-249	65	23	avenues	avenue	NOUN
cril-249	65	24	to	to	PART
cril-249	65	25	travel	travel	VERB
cril-249	65	26	down	down	ADP
cril-249	65	27	for	for	ADP
cril-249	65	28	system	system	NOUN
cril-249	65	29	improvement	improvement	NOUN
cril-249	65	30	.	.	PUNCT
cril-249	66	1	4	4	X
cril-249	66	2	.	.	NOUN
cril-249	66	3	letter	letter	NOUN
cril-249	66	4	to	to	PART
cril-249	66	5	sound	sound	VERB
cril-249	66	6	rules	rule	NOUN
cril-249	66	7	4.1	4.1	NUM
cril-249	66	8	.	.	PUNCT
cril-249	67	1	training	training	NOUN
cril-249	67	2	data	datum	NOUN
cril-249	67	3	our	our	PRON
cril-249	67	4	letter	letter	NOUN
cril-249	67	5	to	to	PART
cril-249	67	6	sound	sound	VERB
cril-249	67	7	output	output	NOUN
cril-249	67	8	systems	system	NOUN
cril-249	67	9	for	for	ADP
cril-249	67	10	arabic	arabic	ADJ
cril-249	67	11	and	and	CCONJ
cril-249	67	12	russian	russian	PROPN
cril-249	67	13	are	be	AUX
cril-249	67	14	built	build	VERB
cril-249	67	15	upon	upon	SCONJ
cril-249	67	16	the	the	DET
cril-249	67	17	integration	integration	NOUN
cril-249	67	18	of	of	ADP
cril-249	67	19	two	two	NUM
cril-249	67	20	separate	separate	ADJ
cril-249	67	21	resources	resource	NOUN
cril-249	67	22	.	.	PUNCT
cril-249	68	1	first	first	ADV
cril-249	68	2	our	our	PRON
cril-249	68	3	lexicon	lexicon	NOUN
cril-249	68	4	is	be	AUX
cril-249	68	5	structured	structure	VERB
cril-249	68	6	in	in	ADP
cril-249	68	7	the	the	DET
cril-249	68	8	same	same	ADJ
cril-249	68	9	way	way	NOUN
cril-249	68	10	as	as	SCONJ
cril-249	68	11	the	the	DET
cril-249	68	12	cmu	cmu	PROPN
cril-249	68	13	pronunciation	pronunciation	NOUN
cril-249	68	14	dictionary	dictionary	NOUN
cril-249	68	15	reformatted	reformatte	VERB
cril-249	68	16	to	to	ADP
cril-249	68	17	sphinx	sphinx	NOUN
cril-249	68	18	format	format	NOUN
cril-249	68	19	(	(	PUNCT
cril-249	68	20	cmu	cmu	NOUN
cril-249	68	21	sphinx	sphinx	PROPN
cril-249	68	22	)	)	PUNCT
cril-249	68	23	.	.	PUNCT
cril-249	69	1	second	second	ADV
cril-249	69	2	we	we	PRON
cril-249	69	3	have	have	AUX
cril-249	69	4	used	use	VERB
cril-249	69	5	the	the	DET
cril-249	69	6	sonic	sonic	ADJ
cril-249	69	7	(	(	PUNCT
cril-249	69	8	pellom	pellom	NOUN
cril-249	69	9	,	,	PUNCT
cril-249	69	10	2003	2003	NUM
cril-249	69	11	)	)	PUNCT
cril-249	69	12	decision	decision	NOUN
cril-249	69	13	tree	tree	NOUN
cril-249	69	14	software	software	NOUN
cril-249	69	15	to	to	PART
cril-249	69	16	train	train	VERB
cril-249	69	17	our	our	PRON
cril-249	69	18	the	the	DET
cril-249	69	19	letter	letter	NOUN
cril-249	69	20	-	-	PUNCT
cril-249	69	21	tosound	tosound	NOUN
cril-249	69	22	rules	rule	NOUN
cril-249	69	23	on	on	ADP
cril-249	69	24	producing	produce	VERB
cril-249	69	25	the	the	DET
cril-249	69	26	correct	correct	ADJ
cril-249	69	27	output	output	NOUN
cril-249	69	28	consisting	consist	VERB
cril-249	69	29	of	of	ADP
cril-249	69	30	a	a	DET
cril-249	69	31	word	word	NOUN
cril-249	69	32	and	and	CCONJ
cril-249	69	33	it	it	PRON
cril-249	69	34	's	be	AUX
cril-249	69	35	arpabet	arpabet	ADJ
cril-249	69	36	transcription	transcription	NOUN
cril-249	69	37	.	.	PUNCT
cril-249	70	1	given	give	VERB
cril-249	70	2	the	the	DET
cril-249	70	3	absence	absence	NOUN
cril-249	70	4	of	of	ADP
cril-249	70	5	a	a	DET
cril-249	70	6	proper	proper	ADJ
cril-249	70	7	corpus	corpus	NOUN
cril-249	70	8	of	of	ADP
cril-249	70	9	arabic	arabic	ADJ
cril-249	70	10	and	and	CCONJ
cril-249	70	11	russian	russian	ADJ
cril-249	70	12	words	word	NOUN
cril-249	70	13	transcribed	transcribe	VERB
cril-249	70	14	in	in	ADP
cril-249	70	15	this	this	DET
cril-249	70	16	manner	manner	NOUN
cril-249	70	17	,	,	PUNCT
cril-249	70	18	we	we	PRON
cril-249	70	19	once	once	ADV
cril-249	70	20	again	again	ADV
cril-249	70	21	scraped	scrape	VERB
cril-249	70	22	the	the	DET
cril-249	70	23	web	web	NOUN
cril-249	70	24	for	for	ADP
cril-249	70	25	data	datum	NOUN
cril-249	70	26	.	.	PUNCT
cril-249	71	1	after	after	ADP
cril-249	71	2	collecting	collect	VERB
cril-249	71	3	transliterated	transliterated	ADJ
cril-249	71	4	words	word	NOUN
cril-249	71	5	from	from	ADP
cril-249	71	6	various	various	ADJ
cril-249	71	7	resources	resource	NOUN
cril-249	71	8	theses	theses	X
cril-249	71	9	words	word	NOUN
cril-249	71	10	were	be	AUX
cril-249	71	11	hand	hand	NOUN
cril-249	71	12	-	-	PUNCT
cril-249	71	13	transcribed	transcribe	VERB
cril-249	71	14	into	into	ADP
cril-249	71	15	arpabet	arpabet	ADJ
cril-249	71	16	phonemic	phonemic	ADJ
cril-249	71	17	output	output	NOUN
cril-249	71	18	by	by	ADP
cril-249	71	19	a	a	DET
cril-249	71	20	team	team	NOUN
cril-249	71	21	of	of	ADP
cril-249	71	22	graduate	graduate	NOUN
cril-249	71	23	linguists	linguist	NOUN
cril-249	71	24	.	.	PUNCT
cril-249	72	1	in	in	ADP
cril-249	72	2	all	all	PRON
cril-249	72	3	we	we	PRON
cril-249	72	4	built	build	VERB
cril-249	72	5	two	two	NUM
cril-249	72	6	corpora	corpus	NOUN
cril-249	72	7	of	of	ADP
cril-249	72	8	transliterated	transliterated	ADJ
cril-249	72	9	words	word	NOUN
cril-249	72	10	,	,	PUNCT
cril-249	72	11	844	844	NUM
cril-249	72	12	russian	russian	ADJ
cril-249	72	13	words	word	NOUN
cril-249	72	14	and	and	CCONJ
cril-249	72	15	582	582	NUM
cril-249	72	16	arabic	arabic	ADJ
cril-249	72	17	words	word	NOUN
cril-249	72	18	.	.	PUNCT
cril-249	73	1	this	this	DET
cril-249	73	2	corpus	corpus	NOUN
cril-249	73	3	is	be	AUX
cril-249	73	4	small	small	ADJ
cril-249	73	5	,	,	PUNCT
cril-249	73	6	but	but	CCONJ
cril-249	73	7	since	since	SCONJ
cril-249	73	8	the	the	DET
cril-249	73	9	words	word	NOUN
cril-249	73	10	were	be	AUX
cril-249	73	11	hand	hand	NOUN
cril-249	73	12	-	-	PUNCT
cril-249	73	13	transcribed	transcribe	VERB
cril-249	73	14	,	,	PUNCT
cril-249	73	15	its	its	PRON
cril-249	73	16	data	data	NOUN
cril-249	73	17	is	be	AUX
cril-249	73	18	at	at	ADP
cril-249	73	19	least	least	ADJ
cril-249	73	20	reliable	reliable	ADJ
cril-249	73	21	.	.	PUNCT
cril-249	74	1	both	both	DET
cril-249	74	2	transcription	transcription	NOUN
cril-249	74	3	time	time	NOUN
cril-249	74	4	and	and	CCONJ
cril-249	74	5	the	the	DET
cril-249	74	6	previous	previous	ADJ
cril-249	74	7	shortage	shortage	NOUN
cril-249	74	8	of	of	ADP
cril-249	74	9	appropriately	appropriately	ADV
cril-249	74	10	transliterated	transliterate	VERB
cril-249	74	11	words	word	NOUN
cril-249	74	12	contributed	contribute	VERB
cril-249	74	13	to	to	ADP
cril-249	74	14	the	the	DET
cril-249	74	15	decision	decision	NOUN
cril-249	74	16	to	to	PART
cril-249	74	17	use	use	VERB
cril-249	74	18	a	a	DET
cril-249	74	19	small	small	ADJ
cril-249	74	20	but	but	CCONJ
cril-249	74	21	reliable	reliable	ADJ
cril-249	74	22	data	datum	NOUN
cril-249	74	23	set	set	VERB
cril-249	74	24	.	.	PUNCT
cril-249	75	1	4.2	4.2	NUM
cril-249	75	2	.	.	PUNCT
cril-249	76	1	training	train	VERB
cril-249	76	2	algorithm	algorithm	NOUN
cril-249	76	3	for	for	ADP
cril-249	76	4	our	our	PRON
cril-249	76	5	decision	decision	NOUN
cril-249	76	6	tree	tree	NOUN
cril-249	76	7	training	training	NOUN
cril-249	76	8	algorithm	algorithm	NOUN
cril-249	76	9	we	we	PRON
cril-249	76	10	used	use	VERB
cril-249	76	11	borrowed	borrow	VERB
cril-249	76	12	technology	technology	NOUN
cril-249	76	13	from	from	ADP
cril-249	76	14	the	the	DET
cril-249	76	15	sonic	sonic	ADJ
cril-249	76	16	system	system	NOUN
cril-249	76	17	.	.	PUNCT
cril-249	77	1	the	the	DET
cril-249	77	2	sonic	sonic	ADJ
cril-249	77	3	system	system	NOUN
cril-249	77	4	allows	allow	VERB
cril-249	77	5	us	we	PRON
cril-249	77	6	to	to	PART
cril-249	77	7	produce	produce	VERB
cril-249	77	8	letter	letter	NOUN
cril-249	77	9	-	-	PUNCT
cril-249	77	10	to	to	ADP
cril-249	77	11	-	-	PUNCT
cril-249	77	12	sound	sound	NOUN
cril-249	77	13	output	output	NOUN
cril-249	77	14	for	for	ADP
cril-249	77	15	words	word	NOUN
cril-249	77	16	which	which	PRON
cril-249	77	17	we	we	PRON
cril-249	77	18	have	have	AUX
cril-249	77	19	never	never	ADV
cril-249	77	20	seen	see	VERB
cril-249	77	21	before	before	ADV
cril-249	77	22	and	and	CCONJ
cril-249	77	23	which	which	PRON
cril-249	77	24	do	do	AUX
cril-249	77	25	not	not	PART
cril-249	77	26	exist	exist	VERB
cril-249	77	27	in	in	ADP
cril-249	77	28	our	our	PRON
cril-249	77	29	language	language	NOUN
cril-249	77	30	specific	specific	ADJ
cril-249	77	31	training	training	NOUN
cril-249	77	32	corpora	corpora	NOUN
cril-249	77	33	.	.	PUNCT
cril-249	78	1	4	4	NUM
cril-249	78	2	colorado	colorado	NOUN
cril-249	78	3	research	research	NOUN
cril-249	78	4	in	in	ADP
cril-249	78	5	linguistics	linguistic	NOUN
cril-249	78	6	,	,	PUNCT
cril-249	78	7	vol	vol	NOUN
cril-249	78	8	.	.	PROPN
cril-249	78	9	17	17	NUM
cril-249	79	1	[	[	X
cril-249	79	2	2004	2004	NUM
cril-249	79	3	]	]	PUNCT
cril-249	79	4	https://scholar.colorado.edu/cril/vol17/iss1/6	https://scholar.colorado.edu/cril/vol17/iss1/6	NOUN
cril-249	79	5	doi	doi	NOUN
cril-249	79	6	:	:	PUNCT
cril-249	79	7	https://doi.org/10.25810/60mf-fn94	https://doi.org/10.25810/60mf-fn94	ADJ
cril-249	79	8	http://fife.speech.cs.cmu.edu/sphinx/	http://fife.speech.cs.cmu.edu/sphinx/	NUM
cril-249	79	9	language	language	NOUN
cril-249	79	10	identification	identification	NOUN
cril-249	79	11	and	and	CCONJ
cril-249	79	12	language	language	NOUN
cril-249	79	13	specific	specific	ADJ
cril-249	79	14	letter	letter	NOUN
cril-249	79	15	-	-	PUNCT
cril-249	79	16	to	to	ADP
cril-249	79	17	-	-	PUNCT
cril-249	79	18	sound	sound	NOUN
cril-249	79	19	rules	rule	NOUN
cril-249	79	20	5	5	NUM
cril-249	79	21	the	the	DET
cril-249	79	22	algorithm	algorithm	NOUN
cril-249	79	23	works	work	VERB
cril-249	79	24	by	by	ADP
cril-249	79	25	extracting	extract	VERB
cril-249	79	26	feature	feature	NOUN
cril-249	79	27	vectors	vector	NOUN
cril-249	79	28	from	from	ADP
cril-249	79	29	the	the	DET
cril-249	79	30	input	input	NOUN
cril-249	79	31	data	datum	NOUN
cril-249	79	32	consisting	consist	VERB
cril-249	79	33	of	of	ADP
cril-249	79	34	the	the	DET
cril-249	79	35	center	center	NOUN
cril-249	79	36	letter	letter	NOUN
cril-249	79	37	plus	plus	CCONJ
cril-249	79	38	3	3	NUM
cril-249	79	39	letters	letter	NOUN
cril-249	79	40	of	of	ADP
cril-249	79	41	context	context	NOUN
cril-249	79	42	.	.	PUNCT
cril-249	80	1	each	each	DET
cril-249	80	2	output	output	NOUN
cril-249	80	3	phoneme	phoneme	NOUN
cril-249	80	4	is	be	AUX
cril-249	80	5	selected	select	VERB
cril-249	80	6	using	use	VERB
cril-249	80	7	a	a	DET
cril-249	80	8	greatest	great	ADJ
cril-249	80	9	reduction	reduction	NOUN
cril-249	80	10	of	of	ADP
cril-249	80	11	entropy	entropy	NOUN
cril-249	80	12	measure	measure	NOUN
cril-249	80	13	.	.	PUNCT
cril-249	81	1	data	datum	NOUN
cril-249	81	2	preparation	preparation	NOUN
cril-249	81	3	for	for	ADP
cril-249	81	4	using	use	VERB
cril-249	81	5	the	the	DET
cril-249	81	6	sonic	sonic	ADJ
cril-249	81	7	system	system	NOUN
cril-249	81	8	requires	require	VERB
cril-249	81	9	input	input	NOUN
cril-249	81	10	and	and	CCONJ
cril-249	81	11	produced	produce	VERB
cril-249	81	12	output	output	NOUN
cril-249	81	13	as	as	SCONJ
cril-249	81	14	shown	show	VERB
cril-249	81	15	in	in	ADP
cril-249	81	16	example	example	NOUN
cril-249	81	17	2	2	NUM
cril-249	81	18	.	.	NOUN
cril-249	81	19	example	example	NOUN
cril-249	81	20	2	2	NUM
cril-249	81	21	:	:	PUNCT
cril-249	81	22	input	input	NOUN
cril-249	81	23	golova	golova	NOUN
cril-249	81	24	g	g	NOUN
cril-249	81	25	ow	ow	INTJ
cril-249	81	26	l	l	NOUN
cril-249	81	27	ax	ax	NOUN
cril-249	81	28	v	v	ADP
cril-249	81	29	aa	aa	PROPN
cril-249	81	30	output	output	NOUN
cril-249	82	1	k	k	PROPN
cril-249	82	2	aa	aa	NOUN
cril-249	82	3	r	r	PROPN
cril-249	82	4	iy	iy	PROPN
cril-249	82	5	k	k	PROPN
cril-249	82	6	uw	uw	PROPN
cril-249	82	7	(	(	PUNCT
cril-249	82	8	caricu	caricu	PROPN
cril-249	82	9	)	)	PUNCT
cril-249	82	10	5	5	NUM
cril-249	82	11	.	.	PUNCT
cril-249	83	1	results	result	NOUN
cril-249	83	2	/	/	SYM
cril-249	83	3	analysis	analysis	NOUN
cril-249	83	4	analysis	analysis	NOUN
cril-249	83	5	of	of	ADP
cril-249	83	6	the	the	DET
cril-249	83	7	system	system	NOUN
cril-249	83	8	was	be	AUX
cril-249	83	9	done	do	VERB
cril-249	83	10	in	in	ADP
cril-249	83	11	3	3	NUM
cril-249	83	12	distinct	distinct	ADJ
cril-249	83	13	phases	phase	NOUN
cril-249	83	14	–	–	PUNCT
cril-249	83	15	analysis	analysis	NOUN
cril-249	83	16	of	of	ADP
cril-249	83	17	the	the	DET
cril-249	83	18	classifier	classifier	NOUN
cril-249	83	19	,	,	PUNCT
cril-249	83	20	analysis	analysis	NOUN
cril-249	83	21	of	of	ADP
cril-249	83	22	the	the	DET
cril-249	83	23	letter	letter	NOUN
cril-249	83	24	-	-	PUNCT
cril-249	83	25	to	to	ADP
cril-249	83	26	-	-	PUNCT
cril-249	83	27	sound	sound	NOUN
cril-249	83	28	rules	rule	NOUN
cril-249	83	29	,	,	PUNCT
cril-249	83	30	and	and	CCONJ
cril-249	83	31	analysis	analysis	NOUN
cril-249	83	32	of	of	ADP
cril-249	83	33	the	the	DET
cril-249	83	34	complete	complete	ADJ
cril-249	83	35	system	system	NOUN
cril-249	83	36	5.1	5.1	NUM
cril-249	83	37	.	.	PUNCT
cril-249	84	1	phase	phase	NOUN
cril-249	84	2	1	1	NUM
cril-249	84	3	:	:	PUNCT
cril-249	84	4	analysis	analysis	NOUN
cril-249	84	5	of	of	ADP
cril-249	84	6	the	the	DET
cril-249	84	7	classifier	classifier	NOUN
cril-249	84	8	the	the	DET
cril-249	84	9	language	language	NOUN
cril-249	84	10	identification	identification	NOUN
cril-249	84	11	classifier	classifier	NOUN
cril-249	84	12	training	training	NOUN
cril-249	84	13	data	datum	NOUN
cril-249	84	14	was	be	AUX
cril-249	84	15	split	split	VERB
cril-249	84	16	to	to	PART
cril-249	84	17	create	create	VERB
cril-249	84	18	test	test	NOUN
cril-249	84	19	data	datum	NOUN
cril-249	84	20	from	from	ADP
cril-249	84	21	20	20	NUM
cril-249	84	22	%	%	NOUN
cril-249	84	23	of	of	ADP
cril-249	84	24	each	each	PRON
cril-249	84	25	of	of	ADP
cril-249	84	26	the	the	DET
cril-249	84	27	three	three	NUM
cril-249	84	28	language	language	NOUN
cril-249	84	29	word	word	NOUN
cril-249	84	30	lists	list	NOUN
cril-249	84	31	.	.	PUNCT
cril-249	85	1	the	the	DET
cril-249	85	2	classifier	classifier	NOUN
cril-249	85	3	was	be	AUX
cril-249	85	4	then	then	ADV
cril-249	85	5	trained	train	VERB
cril-249	85	6	using	use	VERB
cril-249	85	7	the	the	DET
cril-249	85	8	remaining	remain	VERB
cril-249	85	9	80	80	NUM
cril-249	85	10	%	%	NOUN
cril-249	85	11	of	of	ADP
cril-249	85	12	each	each	DET
cril-249	85	13	list	list	NOUN
cril-249	85	14	.	.	PUNCT
cril-249	86	1	this	this	DET
cril-249	86	2	process	process	NOUN
cril-249	86	3	was	be	AUX
cril-249	86	4	repeated	repeat	VERB
cril-249	86	5	4	4	NUM
cril-249	86	6	times	time	NOUN
cril-249	86	7	with	with	ADP
cril-249	86	8	the	the	DET
cril-249	86	9	training	training	NOUN
cril-249	86	10	data	datum	NOUN
cril-249	86	11	split	split	VERB
cril-249	86	12	differently	differently	ADV
cril-249	86	13	each	each	DET
cril-249	86	14	time	time	NOUN
cril-249	86	15	,	,	PUNCT
cril-249	86	16	providing	provide	VERB
cril-249	86	17	5	5	NUM
cril-249	86	18	different	different	ADJ
cril-249	86	19	overlapping	overlap	VERB
cril-249	86	20	training	training	NOUN
cril-249	86	21	sets	set	NOUN
cril-249	86	22	and	and	CCONJ
cril-249	86	23	test	test	NOUN
cril-249	86	24	sets	set	NOUN
cril-249	86	25	.	.	PUNCT
cril-249	87	1	each	each	PRON
cril-249	87	2	was	be	AUX
cril-249	87	3	analyzed	analyze	VERB
cril-249	87	4	for	for	ADP
cril-249	87	5	precision	precision	NOUN
cril-249	87	6	by	by	ADP
cril-249	87	7	simply	simply	ADV
cril-249	87	8	counting	count	VERB
cril-249	87	9	the	the	DET
cril-249	87	10	number	number	NOUN
cril-249	87	11	of	of	ADP
cril-249	87	12	times	time	NOUN
cril-249	87	13	each	each	DET
cril-249	87	14	classifier	classifier	NOUN
cril-249	87	15	correctly	correctly	ADV
cril-249	87	16	labeled	label	VERB
cril-249	87	17	the	the	DET
cril-249	87	18	words	word	NOUN
cril-249	87	19	in	in	ADP
cril-249	87	20	the	the	DET
cril-249	87	21	test	test	NOUN
cril-249	87	22	sets	set	NOUN
cril-249	87	23	for	for	ADP
cril-249	87	24	each	each	DET
cril-249	87	25	language	language	NOUN
cril-249	87	26	.	.	PUNCT
cril-249	88	1	the	the	DET
cril-249	88	2	classifier	classifier	NOUN
cril-249	88	3	's	's	PART
cril-249	88	4	precision	precision	NOUN
cril-249	88	5	on	on	ADP
cril-249	88	6	the	the	DET
cril-249	88	7	arabic	arabic	ADJ
cril-249	88	8	test	test	NOUN
cril-249	88	9	data	datum	NOUN
cril-249	88	10	sets	set	NOUN
cril-249	88	11	described	describe	VERB
cril-249	88	12	above	above	ADP
cril-249	88	13	ranged	range	VERB
cril-249	88	14	from	from	ADP
cril-249	88	15	.86	.86	NUM
cril-249	88	16	to	to	ADP
cril-249	88	17	1.0	1.0	NUM
cril-249	88	18	with	with	ADP
cril-249	88	19	a	a	DET
cril-249	88	20	mean	mean	NOUN
cril-249	88	21	of	of	ADP
cril-249	88	22	.92	.92	NUM
cril-249	88	23	.	.	PUNCT
cril-249	89	1	on	on	ADP
cril-249	89	2	the	the	DET
cril-249	89	3	russian	russian	ADJ
cril-249	89	4	data	datum	NOUN
cril-249	89	5	precision	precision	NOUN
cril-249	89	6	ranged	range	VERB
cril-249	89	7	from	from	ADP
cril-249	89	8	.79	.79	NUM
cril-249	89	9	to	to	ADP
cril-249	89	10	.83	.83	NUM
cril-249	89	11	with	with	ADP
cril-249	89	12	a	a	DET
cril-249	89	13	mean	mean	NOUN
cril-249	89	14	of	of	ADP
cril-249	89	15	.81	.81	NUM
cril-249	89	16	.	.	PUNCT
cril-249	90	1	the	the	DET
cril-249	90	2	classifier	classifier	NOUN
cril-249	90	3	achieved	achieve	VERB
cril-249	90	4	its	its	PRON
cril-249	90	5	highest	high	ADJ
cril-249	90	6	accuracy	accuracy	NOUN
cril-249	90	7	and	and	CCONJ
cril-249	90	8	greatest	great	ADJ
cril-249	90	9	consistency	consistency	NOUN
cril-249	90	10	on	on	ADP
cril-249	90	11	the	the	DET
cril-249	90	12	english	english	PROPN
cril-249	90	13	data	datum	NOUN
cril-249	90	14	with	with	ADP
cril-249	90	15	a	a	DET
cril-249	90	16	precision	precision	NOUN
cril-249	90	17	ranging	range	VERB
cril-249	90	18	from	from	ADP
cril-249	90	19	.986	.986	NUM
cril-249	90	20	to	to	ADP
cril-249	90	21	.992	.992	NUM
cril-249	90	22	and	and	CCONJ
cril-249	90	23	a	a	DET
cril-249	90	24	mean	mean	NOUN
cril-249	90	25	of	of	ADP
cril-249	90	26	.988	.988	NUM
cril-249	90	27	.	.	PUNCT
cril-249	91	1	5	5	NUM
cril-249	91	2	lewis	lewis	PROPN
cril-249	91	3	et	et	PROPN
cril-249	91	4	al	al	PROPN
cril-249	91	5	.	.	PUNCT
cril-249	91	6	:	:	PUNCT
cril-249	92	1	language	language	NOUN
cril-249	92	2	identification	identification	NOUN
cril-249	92	3	and	and	CCONJ
cril-249	92	4	language	language	NOUN
cril-249	92	5	specific	specific	ADJ
cril-249	92	6	letter	letter	NOUN
cril-249	92	7	-	-	PUNCT
cril-249	92	8	to	to	ADP
cril-249	92	9	-	-	PUNCT
cril-249	92	10	sound	sound	NOUN
cril-249	92	11	rules	rule	NOUN
cril-249	92	12	published	publish	VERB
cril-249	92	13	by	by	ADP
cril-249	92	14	cu	cu	PROPN
cril-249	92	15	scholar	scholar	NOUN
cril-249	92	16	,	,	PUNCT
cril-249	92	17	2004	2004	NUM
cril-249	92	18	colorado	colorado	PROPN
cril-249	92	19	research	research	NOUN
cril-249	92	20	in	in	ADP
cril-249	92	21	linguistics	linguistic	NOUN
cril-249	92	22	,	,	PUNCT
cril-249	92	23	volume	volume	NOUN
cril-249	92	24	17	17	NUM
cril-249	92	25	(	(	PUNCT
cril-249	92	26	2004	2004	NUM
cril-249	92	27	)	)	PUNCT
cril-249	92	28	6	6	NUM
cril-249	92	29	language	language	NOUN
cril-249	92	30	identification	identification	NOUN
cril-249	92	31	performance	performance	NOUN
cril-249	92	32	0.5	0.5	NUM
cril-249	92	33	0.55	0.55	NUM
cril-249	92	34	0.6	0.6	NUM
cril-249	92	35	0.65	0.65	NUM
cril-249	92	36	0.7	0.7	NUM
cril-249	92	37	0.75	0.75	NUM
cril-249	92	38	0.8	0.8	NUM
cril-249	92	39	0.85	0.85	NUM
cril-249	92	40	0.9	0.9	NUM
cril-249	92	41	0.95	0.95	NUM
cril-249	92	42	1	1	NUM
cril-249	92	43	classifier	classifier	NOUN
cril-249	92	44	1	1	NUM
cril-249	92	45	0.998	0.998	NUM
cril-249	92	46	0.988	0.988	NUM
cril-249	92	47	0.804	0.804	NUM
cril-249	92	48	classifier	classifier	NOUN
cril-249	92	49	2	2	NUM
cril-249	92	50	0.87	0.87	NUM
cril-249	92	51	0.987	0.987	NUM
cril-249	92	52	0.833	0.833	NUM
cril-249	92	53	classifier	classifier	NOUN
cril-249	92	54	3	3	NUM
cril-249	92	55	1	1	NUM
cril-249	92	56	0.992	0.992	NUM
cril-249	92	57	0.829	0.829	NUM
cril-249	92	58	classifier	classifier	NOUN
cril-249	92	59	4	4	NUM
cril-249	92	60	0.874	0.874	NUM
cril-249	92	61	0.989	0.989	NUM
cril-249	92	62	0.801	0.801	NUM
cril-249	92	63	classifier	classifier	NOUN
cril-249	92	64	5	5	NUM
cril-249	92	65	0.855	0.855	NUM
cril-249	92	66	0.988	0.988	NUM
cril-249	92	67	0.79	0.79	NUM
cril-249	92	68	mean	mean	NOUN
cril-249	92	69	0.92	0.92	NUM
cril-249	92	70	0.989	0.989	NUM
cril-249	92	71	0.811	0.811	NUM
cril-249	92	72	arabic	arabic	ADJ
cril-249	92	73	english	english	ADJ
cril-249	92	74	russian	russian	ADJ
cril-249	92	75	figure	figure	NOUN
cril-249	92	76	1	1	NUM
cril-249	92	77	.	.	PUNCT
cril-249	92	78	performance	performance	NOUN
cril-249	92	79	of	of	ADP
cril-249	92	80	the	the	DET
cril-249	92	81	classifier	classifier	NOUN
cril-249	92	82	as	as	SCONJ
cril-249	92	83	trained	train	VERB
cril-249	92	84	and	and	CCONJ
cril-249	92	85	tested	test	VERB
cril-249	92	86	on	on	ADP
cril-249	92	87	5	5	NUM
cril-249	92	88	overlapping	overlap	VERB
cril-249	92	89	data	data	NOUN
cril-249	92	90	sets	set	NOUN
cril-249	92	91	(	(	PUNCT
cril-249	92	92	classifiers	classifier	NOUN
cril-249	92	93	1	1	NUM
cril-249	92	94	-	-	SYM
cril-249	92	95	5	5	NUM
cril-249	92	96	)	)	PUNCT
cril-249	92	97	5.2	5.2	NUM
cril-249	92	98	.	.	PUNCT
cril-249	92	99	phase	phase	NOUN
cril-249	92	100	2	2	NUM
cril-249	92	101	:	:	PUNCT
cril-249	92	102	analysis	analysis	NOUN
cril-249	92	103	of	of	ADP
cril-249	92	104	the	the	DET
cril-249	92	105	letter	letter	NOUN
cril-249	92	106	-	-	PUNCT
cril-249	92	107	to	to	ADP
cril-249	92	108	-	-	PUNCT
cril-249	92	109	sound	sound	NOUN
cril-249	92	110	rules	rule	NOUN
cril-249	92	111	of	of	ADP
cril-249	92	112	the	the	DET
cril-249	92	113	hand	hand	NOUN
cril-249	92	114	transcribed	transcribe	VERB
cril-249	92	115	arabic	arabic	ADJ
cril-249	92	116	and	and	CCONJ
cril-249	92	117	russian	russian	ADJ
cril-249	92	118	transliterated	transliterated	ADJ
cril-249	92	119	words	word	NOUN
cril-249	92	120	,	,	PUNCT
cril-249	92	121	10	10	NUM
cril-249	92	122	%	%	NOUN
cril-249	92	123	were	be	AUX
cril-249	92	124	reserved	reserve	VERB
cril-249	92	125	for	for	ADP
cril-249	92	126	testing	test	VERB
cril-249	92	127	the	the	DET
cril-249	92	128	language	language	NOUN
cril-249	92	129	specific	specific	ADJ
cril-249	92	130	letter	letter	NOUN
cril-249	92	131	-	-	PUNCT
cril-249	92	132	to	to	ADP
cril-249	92	133	-	-	PUNCT
cril-249	92	134	sound	sound	NOUN
cril-249	92	135	rule	rule	NOUN
cril-249	92	136	decision	decision	NOUN
cril-249	92	137	trees	tree	NOUN
cril-249	92	138	.	.	PUNCT
cril-249	93	1	performance	performance	NOUN
cril-249	93	2	of	of	ADP
cril-249	93	3	the	the	DET
cril-249	93	4	language	language	NOUN
cril-249	93	5	specific	specific	ADJ
cril-249	93	6	decision	decision	NOUN
cril-249	93	7	trees	tree	NOUN
cril-249	93	8	was	be	AUX
cril-249	93	9	compared	compare	VERB
cril-249	93	10	against	against	ADP
cril-249	93	11	the	the	DET
cril-249	93	12	performance	performance	NOUN
cril-249	93	13	of	of	ADP
cril-249	93	14	a	a	DET
cril-249	93	15	decision	decision	NOUN
cril-249	93	16	tree	tree	NOUN
cril-249	93	17	trained	train	VERB
cril-249	93	18	on	on	ADP
cril-249	93	19	the	the	DET
cril-249	93	20	cmu	cmu	PROPN
cril-249	93	21	lexicon	lexicon	NOUN
cril-249	93	22	as	as	ADP
cril-249	93	23	a	a	DET
cril-249	93	24	baseline	baseline	NOUN
cril-249	93	25	.	.	PUNCT
cril-249	94	1	accuracy	accuracy	NOUN
cril-249	94	2	was	be	AUX
cril-249	94	3	determined	determine	VERB
cril-249	94	4	by	by	ADP
cril-249	94	5	counting	count	VERB
cril-249	94	6	the	the	DET
cril-249	94	7	phones	phone	NOUN
cril-249	94	8	that	that	PRON
cril-249	94	9	matched	match	VERB
cril-249	94	10	between	between	ADP
cril-249	94	11	the	the	DET
cril-249	94	12	hand	hand	NOUN
cril-249	94	13	transcription	transcription	NOUN
cril-249	94	14	and	and	CCONJ
cril-249	94	15	the	the	DET
cril-249	94	16	decision	decision	NOUN
cril-249	94	17	tree	tree	NOUN
cril-249	94	18	transcription	transcription	NOUN
cril-249	94	19	.	.	PUNCT
cril-249	95	1	the	the	DET
cril-249	95	2	russian	russian	ADJ
cril-249	95	3	decision	decision	NOUN
cril-249	95	4	tree	tree	NOUN
cril-249	95	5	performed	perform	VERB
cril-249	95	6	at	at	ADP
cril-249	95	7	a	a	DET
cril-249	95	8	precision	precision	NOUN
cril-249	95	9	of	of	ADP
cril-249	95	10	.93	.93	NUM
cril-249	95	11	(	(	PUNCT
cril-249	95	12	error	error	NOUN
cril-249	95	13	rate	rate	NOUN
cril-249	95	14	.07	.07	NUM
cril-249	95	15	)	)	PUNCT
cril-249	95	16	on	on	ADP
cril-249	95	17	the	the	DET
cril-249	95	18	russian	russian	ADJ
cril-249	95	19	test	test	NOUN
cril-249	95	20	set	set	VERB
cril-249	95	21	.	.	PUNCT
cril-249	96	1	the	the	DET
cril-249	96	2	baseline	baseline	ADJ
cril-249	96	3	decision	decision	NOUN
cril-249	96	4	tree	tree	NOUN
cril-249	96	5	performed	perform	VERB
cril-249	96	6	at	at	ADP
cril-249	96	7	a	a	DET
cril-249	96	8	precision	precision	NOUN
cril-249	96	9	of	of	ADP
cril-249	96	10	.66	.66	NUM
cril-249	96	11	(	(	PUNCT
cril-249	96	12	error	error	NOUN
cril-249	96	13	rate	rate	NOUN
cril-249	96	14	.33	.33	NUM
cril-249	96	15	)	)	PUNCT
cril-249	96	16	on	on	ADP
cril-249	96	17	the	the	DET
cril-249	96	18	0.5	0.5	NUM
cril-249	96	19	0.7	0.7	NUM
cril-249	96	20	0.9	0.9	NUM
cril-249	96	21	1.1	1.1	NUM
cril-249	96	22	letter	letter	NOUN
cril-249	96	23	-	-	PUNCT
cril-249	96	24	to	to	ADP
cril-249	96	25	-	-	PUNCT
cril-249	96	26	sound	sound	NOUN
cril-249	96	27	performance	performance	NOUN
cril-249	96	28	baseline	baseline	NOUN
cril-249	96	29	(	(	PUNCT
cril-249	96	30	cmu	cmu	PROPN
cril-249	96	31	)	)	PUNCT
cril-249	96	32	0.711	0.711	NUM
cril-249	96	33	0.662	0.662	NUM
cril-249	96	34	language	language	NOUN
cril-249	96	35	specific	specific	ADJ
cril-249	96	36	0.877	0.877	NUM
cril-249	96	37	0.928	0.928	NUM
cril-249	96	38	arabic	arabic	ADJ
cril-249	96	39	russian	russian	ADJ
cril-249	96	40	figure	figure	NOUN
cril-249	96	41	2	2	NUM
cril-249	96	42	.	.	PUNCT
cril-249	96	43	performance	performance	NOUN
cril-249	96	44	of	of	ADP
cril-249	96	45	language	language	NOUN
cril-249	96	46	specific	specific	ADJ
cril-249	96	47	letter	letter	NOUN
cril-249	96	48	to	to	PART
cril-249	96	49	sound	sound	VERB
cril-249	96	50	rules	rule	NOUN
cril-249	96	51	on	on	ADP
cril-249	96	52	language	language	NOUN
cril-249	96	53	specific	specific	ADJ
cril-249	96	54	data	datum	NOUN
cril-249	96	55	as	as	SCONJ
cril-249	96	56	compared	compare	VERB
cril-249	96	57	to	to	PART
cril-249	96	58	baseline	baseline	VERB
cril-249	96	59	6	6	NUM
cril-249	96	60	colorado	colorado	NOUN
cril-249	96	61	research	research	NOUN
cril-249	96	62	in	in	ADP
cril-249	96	63	linguistics	linguistic	NOUN
cril-249	96	64	,	,	PUNCT
cril-249	96	65	vol	vol	NOUN
cril-249	96	66	.	.	PROPN
cril-249	96	67	17	17	NUM
cril-249	97	1	[	[	X
cril-249	97	2	2004	2004	NUM
cril-249	97	3	]	]	PUNCT
cril-249	97	4	https://scholar.colorado.edu/cril/vol17/iss1/6	https://scholar.colorado.edu/cril/vol17/iss1/6	NOUN
cril-249	97	5	doi	doi	NOUN
cril-249	97	6	:	:	PUNCT
cril-249	97	7	https://doi.org/10.25810/60mf-fn94	https://doi.org/10.25810/60mf-fn94	ADJ
cril-249	97	8	language	language	NOUN
cril-249	97	9	identification	identification	NOUN
cril-249	97	10	and	and	CCONJ
cril-249	97	11	language	language	NOUN
cril-249	97	12	specific	specific	ADJ
cril-249	97	13	letter	letter	NOUN
cril-249	97	14	-	-	PUNCT
cril-249	97	15	to	to	ADP
cril-249	97	16	-	-	PUNCT
cril-249	97	17	sound	sound	NOUN
cril-249	97	18	rules	rule	NOUN
cril-249	97	19	7	7	NUM
cril-249	97	20	same	same	ADJ
cril-249	97	21	test	test	NOUN
cril-249	97	22	set	set	NOUN
cril-249	97	23	,	,	PUNCT
cril-249	97	24	showing	show	VERB
cril-249	97	25	a	a	DET
cril-249	97	26	79	79	NUM
cril-249	97	27	percent	percent	NOUN
cril-249	97	28	reduction	reduction	NOUN
cril-249	97	29	in	in	ADP
cril-249	97	30	error	error	NOUN
cril-249	97	31	rate	rate	NOUN
cril-249	97	32	by	by	ADP
cril-249	97	33	using	use	VERB
cril-249	97	34	the	the	DET
cril-249	97	35	russian	russian	ADJ
cril-249	97	36	decision	decision	NOUN
cril-249	97	37	tree	tree	NOUN
cril-249	97	38	.	.	PUNCT
cril-249	98	1	the	the	DET
cril-249	98	2	arabic	arabic	ADJ
cril-249	98	3	decision	decision	NOUN
cril-249	98	4	tree	tree	NOUN
cril-249	98	5	performed	perform	VERB
cril-249	98	6	at	at	ADP
cril-249	98	7	a	a	DET
cril-249	98	8	precision	precision	NOUN
cril-249	98	9	of	of	ADP
cril-249	98	10	.88	.88	NUM
cril-249	98	11	(	(	PUNCT
cril-249	98	12	error	error	NOUN
cril-249	98	13	rate	rate	NOUN
cril-249	98	14	.12	.12	NUM
cril-249	98	15	)	)	PUNCT
cril-249	98	16	on	on	ADP
cril-249	98	17	the	the	DET
cril-249	98	18	arabic	arabic	ADJ
cril-249	98	19	test	test	NOUN
cril-249	98	20	set	set	VERB
cril-249	98	21	.	.	PUNCT
cril-249	99	1	the	the	DET
cril-249	99	2	baseline	baseline	ADJ
cril-249	99	3	decision	decision	NOUN
cril-249	99	4	tree	tree	NOUN
cril-249	99	5	performed	perform	VERB
cril-249	99	6	at	at	ADP
cril-249	99	7	a	a	DET
cril-249	99	8	precision	precision	NOUN
cril-249	99	9	of	of	ADP
cril-249	99	10	.71	.71	NUM
cril-249	99	11	(	(	PUNCT
cril-249	99	12	error	error	NOUN
cril-249	99	13	rate	rate	NOUN
cril-249	99	14	.29	.29	NUM
cril-249	99	15	)	)	PUNCT
cril-249	99	16	on	on	ADP
cril-249	99	17	the	the	DET
cril-249	99	18	same	same	ADJ
cril-249	99	19	test	test	NOUN
cril-249	99	20	set	set	NOUN
cril-249	99	21	,	,	PUNCT
cril-249	99	22	showing	show	VERB
cril-249	99	23	a	a	DET
cril-249	99	24	57	57	NUM
cril-249	99	25	percent	percent	NOUN
cril-249	99	26	reduction	reduction	NOUN
cril-249	99	27	in	in	ADP
cril-249	99	28	error	error	NOUN
cril-249	99	29	rate	rate	NOUN
cril-249	99	30	by	by	ADP
cril-249	99	31	using	use	VERB
cril-249	99	32	the	the	DET
cril-249	99	33	arabic	arabic	ADJ
cril-249	99	34	decision	decision	NOUN
cril-249	99	35	tree	tree	NOUN
cril-249	99	36	.	.	PUNCT
cril-249	100	1	5.3	5.3	NUM
cril-249	100	2	.	.	PUNCT
cril-249	100	3	phase	phase	NOUN
cril-249	100	4	3	3	NUM
cril-249	100	5	:	:	PUNCT
cril-249	100	6	analysis	analysis	NOUN
cril-249	100	7	of	of	ADP
cril-249	100	8	the	the	DET
cril-249	100	9	complete	complete	ADJ
cril-249	100	10	system	system	NOUN
cril-249	100	11	to	to	PART
cril-249	100	12	test	test	VERB
cril-249	100	13	the	the	DET
cril-249	100	14	complete	complete	ADJ
cril-249	100	15	system	system	NOUN
cril-249	100	16	,	,	PUNCT
cril-249	100	17	we	we	PRON
cril-249	100	18	isolated	isolate	VERB
cril-249	100	19	50	50	NUM
cril-249	100	20	words	word	NOUN
cril-249	100	21	from	from	ADP
cril-249	100	22	each	each	DET
cril-249	100	23	hand	hand	NOUN
cril-249	100	24	transcribed	transcribed	NOUN
cril-249	100	25	test	test	NOUN
cril-249	100	26	set	set	VERB
cril-249	100	27	.	.	PUNCT
cril-249	101	1	we	we	PRON
cril-249	101	2	then	then	ADV
cril-249	101	3	isolated	isolate	VERB
cril-249	101	4	50	50	NUM
cril-249	101	5	english	english	ADJ
cril-249	101	6	words	word	NOUN
cril-249	101	7	from	from	ADP
cril-249	101	8	the	the	DET
cril-249	101	9	cmu	cmu	PROPN
cril-249	101	10	lexicon	lexicon	NOUN
cril-249	101	11	,	,	PUNCT
cril-249	101	12	and	and	CCONJ
cril-249	101	13	combined	combine	VERB
cril-249	101	14	the	the	DET
cril-249	101	15	three	three	NUM
cril-249	101	16	sets	set	NOUN
cril-249	101	17	,	,	PUNCT
cril-249	101	18	making	make	VERB
cril-249	101	19	150	150	NUM
cril-249	101	20	transcribed	transcribe	VERB
cril-249	101	21	words	word	NOUN
cril-249	101	22	,	,	PUNCT
cril-249	101	23	50	50	NUM
cril-249	101	24	in	in	ADP
cril-249	101	25	each	each	DET
cril-249	101	26	language	language	NOUN
cril-249	101	27	.	.	PUNCT
cril-249	102	1	these	these	DET
cril-249	102	2	words	word	NOUN
cril-249	102	3	were	be	AUX
cril-249	102	4	classified	classify	VERB
cril-249	102	5	using	use	VERB
cril-249	102	6	the	the	DET
cril-249	102	7	language	language	NOUN
cril-249	102	8	identification	identification	NOUN
cril-249	102	9	classifier	classifier	NOUN
cril-249	102	10	and	and	CCONJ
cril-249	102	11	then	then	ADV
cril-249	102	12	transcribed	transcribe	VERB
cril-249	102	13	using	use	VERB
cril-249	102	14	appropriate	appropriate	ADJ
cril-249	102	15	decision	decision	NOUN
cril-249	102	16	trees	tree	NOUN
cril-249	102	17	based	base	VERB
cril-249	102	18	on	on	ADP
cril-249	102	19	the	the	DET
cril-249	102	20	classifier	classifier	NOUN
cril-249	102	21	labels	label	NOUN
cril-249	102	22	.	.	PUNCT
cril-249	103	1	counts	count	NOUN
cril-249	103	2	were	be	AUX
cril-249	103	3	then	then	ADV
cril-249	103	4	made	make	VERB
cril-249	103	5	of	of	ADP
cril-249	103	6	the	the	DET
cril-249	103	7	phones	phone	NOUN
cril-249	103	8	in	in	ADP
cril-249	103	9	the	the	DET
cril-249	103	10	system	system	NOUN
cril-249	103	11	's	's	PART
cril-249	103	12	transcription	transcription	NOUN
cril-249	103	13	that	that	PRON
cril-249	103	14	matched	match	VERB
cril-249	103	15	the	the	DET
cril-249	103	16	hand	hand	NOUN
cril-249	103	17	transcription	transcription	NOUN
cril-249	103	18	and	and	CCONJ
cril-249	103	19	lexicon	lexicon	NOUN
cril-249	103	20	transcription	transcription	NOUN
cril-249	103	21	.	.	PUNCT
cril-249	104	1	on	on	ADP
cril-249	104	2	this	this	DET
cril-249	104	3	data	data	NOUN
cril-249	104	4	,	,	PUNCT
cril-249	104	5	the	the	DET
cril-249	104	6	system	system	NOUN
cril-249	104	7	achieved	achieve	VERB
cril-249	104	8	a	a	DET
cril-249	104	9	precision	precision	NOUN
cril-249	104	10	of	of	ADP
cril-249	104	11	.89	.89	NUM
cril-249	104	12	.	.	PUNCT
cril-249	105	1	the	the	DET
cril-249	105	2	decision	decision	NOUN
cril-249	105	3	tree	tree	NOUN
cril-249	105	4	trained	train	VERB
cril-249	105	5	on	on	ADP
cril-249	105	6	the	the	DET
cril-249	105	7	cmu	cmu	PROPN
cril-249	105	8	lexicon	lexicon	NOUN
cril-249	105	9	performed	perform	VERB
cril-249	105	10	with	with	ADP
cril-249	105	11	a	a	DET
cril-249	105	12	precision	precision	NOUN
cril-249	105	13	of	of	ADP
cril-249	105	14	.80	.80	NUM
cril-249	105	15	on	on	ADP
cril-249	105	16	the	the	DET
cril-249	105	17	same	same	ADJ
cril-249	105	18	data	datum	NOUN
cril-249	105	19	,	,	PUNCT
cril-249	105	20	showing	show	VERB
cril-249	105	21	a	a	DET
cril-249	105	22	.46	.46	NUM
cril-249	105	23	reduction	reduction	NOUN
cril-249	105	24	in	in	ADP
cril-249	105	25	error	error	NOUN
cril-249	105	26	rate	rate	NOUN
cril-249	105	27	by	by	ADP
cril-249	105	28	using	use	VERB
cril-249	105	29	the	the	DET
cril-249	105	30	new	new	ADJ
cril-249	105	31	system	system	NOUN
cril-249	105	32	.	.	PUNCT
cril-249	106	1	0.5	0.5	NUM
cril-249	106	2	0.7	0.7	NUM
cril-249	106	3	0.9	0.9	NUM
cril-249	106	4	complete	complete	ADJ
cril-249	106	5	system	system	NOUN
cril-249	106	6	performance	performance	NOUN
cril-249	106	7	baseline	baseline	NOUN
cril-249	106	8	(	(	PUNCT
cril-249	106	9	cmu	cmu	PROPN
cril-249	106	10	)	)	PUNCT
cril-249	106	11	0.801	0.801	NUM
cril-249	106	12	ear	ear	VERB
cril-249	106	13	0.892	0.892	NUM
cril-249	106	14	combined	combine	VERB
cril-249	106	15	language	language	NOUN
cril-249	106	16	score	score	NOUN
cril-249	106	17	figure	figure	NOUN
cril-249	106	18	3	3	NUM
cril-249	106	19	.	.	PUNCT
cril-249	106	20	performance	performance	NOUN
cril-249	106	21	of	of	ADP
cril-249	106	22	the	the	DET
cril-249	106	23	language	language	NOUN
cril-249	106	24	identification	identification	NOUN
cril-249	106	25	classifier	classifier	NOUN
cril-249	106	26	and	and	CCONJ
cril-249	106	27	the	the	DET
cril-249	106	28	language	language	NOUN
cril-249	106	29	specific	specific	ADJ
cril-249	106	30	letter	letter	NOUN
cril-249	106	31	-	-	PUNCT
cril-249	106	32	to	to	ADP
cril-249	106	33	-	-	PUNCT
cril-249	106	34	sound	sound	NOUN
cril-249	106	35	rules	rule	NOUN
cril-249	106	36	combined	combine	VERB
cril-249	106	37	on	on	ADP
cril-249	106	38	50	50	NUM
cril-249	106	39	english	english	ADJ
cril-249	106	40	,	,	PUNCT
cril-249	106	41	russian	russian	ADJ
cril-249	106	42	,	,	PUNCT
cril-249	106	43	and	and	CCONJ
cril-249	106	44	arabic	arabic	ADJ
cril-249	106	45	words	word	NOUN
cril-249	106	46	6	6	NUM
cril-249	106	47	.	.	PUNCT
cril-249	107	1	future	future	ADJ
cril-249	107	2	work	work	NOUN
cril-249	107	3	and	and	CCONJ
cril-249	107	4	conclusions	conclusion	NOUN
cril-249	107	5	our	our	PRON
cril-249	107	6	results	result	NOUN
cril-249	107	7	are	be	AUX
cril-249	107	8	quite	quite	ADV
cril-249	107	9	promising	promising	ADJ
cril-249	107	10	.	.	PUNCT
cril-249	108	1	we	we	PRON
cril-249	108	2	have	have	AUX
cril-249	108	3	shown	show	VERB
cril-249	108	4	that	that	SCONJ
cril-249	108	5	we	we	PRON
cril-249	108	6	can	can	AUX
cril-249	108	7	significantly	significantly	ADV
cril-249	108	8	improve	improve	VERB
cril-249	108	9	automatic	automatic	ADJ
cril-249	108	10	transcription	transcription	NOUN
cril-249	108	11	by	by	ADP
cril-249	108	12	integrating	integrate	VERB
cril-249	108	13	an	an	DET
cril-249	108	14	n	n	CCONJ
cril-249	108	15	-	-	PUNCT
cril-249	108	16	gram	gram	NOUN
cril-249	108	17	language	language	NOUN
cril-249	108	18	classifier	classifier	NOUN
cril-249	108	19	and	and	CCONJ
cril-249	108	20	language	language	NOUN
cril-249	108	21	specific	specific	ADJ
cril-249	108	22	transcription	transcription	NOUN
cril-249	108	23	decision	decision	NOUN
cril-249	108	24	trees	tree	NOUN
cril-249	108	25	.	.	PUNCT
cril-249	109	1	these	these	DET
cril-249	109	2	promising	promising	ADJ
cril-249	109	3	results	result	NOUN
cril-249	109	4	warrant	warrant	VERB
cril-249	109	5	further	further	ADJ
cril-249	109	6	experimentation	experimentation	NOUN
cril-249	109	7	.	.	PUNCT
cril-249	110	1	there	there	PRON
cril-249	110	2	is	be	VERB
cril-249	110	3	a	a	DET
cril-249	110	4	wide	wide	ADJ
cril-249	110	5	range	range	NOUN
cril-249	110	6	of	of	ADP
cril-249	110	7	machine	machine	NOUN
cril-249	110	8	learning	learn	VERB
cril-249	110	9	techniques	technique	NOUN
cril-249	110	10	which	which	PRON
cril-249	110	11	we	we	PRON
cril-249	110	12	would	would	AUX
cril-249	110	13	like	like	VERB
cril-249	110	14	to	to	PART
cril-249	110	15	implement	implement	VERB
cril-249	110	16	in	in	ADP
cril-249	110	17	the	the	DET
cril-249	110	18	future	future	NOUN
cril-249	110	19	as	as	ADP
cril-249	110	20	an	an	DET
cril-249	110	21	alternative	alternative	NOUN
cril-249	110	22	to	to	ADP
cril-249	110	23	using	use	VERB
cril-249	110	24	simple	simple	ADJ
cril-249	110	25	n	n	CCONJ
cril-249	110	26	-	-	PUNCT
cril-249	110	27	grams	gram	NOUN
cril-249	110	28	for	for	ADP
cril-249	110	29	the	the	DET
cril-249	110	30	first	first	ADJ
cril-249	110	31	tier	tier	NOUN
cril-249	110	32	of	of	ADP
cril-249	110	33	our	our	PRON
cril-249	110	34	classifier	classifier	NOUN
cril-249	110	35	.	.	PUNCT
cril-249	111	1	we	we	PRON
cril-249	111	2	would	would	AUX
cril-249	111	3	like	like	VERB
cril-249	111	4	to	to	PART
cril-249	111	5	thank	thank	VERB
cril-249	111	6	fellow	fellow	PROPN
cril-249	111	7	researcher	researcher	PROPN
cril-249	111	8	dan	dan	PROPN
cril-249	111	9	cer	cer	PROPN
cril-249	111	10	for	for	ADP
cril-249	111	11	his	his	PRON
cril-249	111	12	suggestion	suggestion	NOUN
cril-249	111	13	that	that	SCONJ
cril-249	111	14	an	an	DET
cril-249	111	15	unsupervised	unsupervised	ADJ
cril-249	111	16	machine	machine	NOUN
cril-249	111	17	learning	learn	VERB
cril-249	111	18	approach	approach	NOUN
cril-249	111	19	could	could	AUX
cril-249	111	20	be	be	AUX
cril-249	111	21	applied	apply	VERB
cril-249	111	22	to	to	ADP
cril-249	111	23	this	this	DET
cril-249	111	24	classification	classification	NOUN
cril-249	111	25	task	task	NOUN
cril-249	111	26	.	.	PUNCT
cril-249	112	1	preliminary	preliminary	ADJ
cril-249	112	2	investigations	investigation	NOUN
cril-249	112	3	imply	imply	VERB
cril-249	112	4	that	that	SCONJ
cril-249	112	5	unsupervised	unsupervised	ADJ
cril-249	112	6	ml	ml	NOUN
cril-249	112	7	is	be	AUX
cril-249	112	8	promising	promise	VERB
cril-249	112	9	.	.	PUNCT
cril-249	113	1	7	7	NUM
cril-249	113	2	lewis	lewis	PROPN
cril-249	113	3	et	et	PROPN
cril-249	113	4	al	al	PROPN
cril-249	113	5	.	.	PUNCT
cril-249	113	6	:	:	PUNCT
cril-249	113	7	language	language	NOUN
cril-249	113	8	identification	identification	NOUN
cril-249	113	9	and	and	CCONJ
cril-249	113	10	language	language	NOUN
cril-249	113	11	specific	specific	ADJ
cril-249	113	12	letter	letter	NOUN
cril-249	113	13	-	-	PUNCT
cril-249	113	14	to	to	ADP
cril-249	113	15	-	-	PUNCT
cril-249	113	16	sound	sound	NOUN
cril-249	113	17	rules	rule	NOUN
cril-249	113	18	published	publish	VERB
cril-249	113	19	by	by	ADP
cril-249	113	20	cu	cu	PROPN
cril-249	113	21	scholar	scholar	NOUN
cril-249	113	22	,	,	PUNCT
cril-249	113	23	2004	2004	NUM
cril-249	113	24	colorado	colorado	PROPN
cril-249	113	25	research	research	NOUN
cril-249	113	26	in	in	ADP
cril-249	113	27	linguistics	linguistic	NOUN
cril-249	113	28	,	,	PUNCT
cril-249	113	29	volume	volume	NOUN
cril-249	113	30	17	17	NUM
cril-249	113	31	(	(	PUNCT
cril-249	113	32	2004	2004	NUM
cril-249	113	33	)	)	PUNCT
cril-249	113	34	8	8	NUM
cril-249	113	35	it	it	PRON
cril-249	113	36	is	be	AUX
cril-249	113	37	likely	likely	ADJ
cril-249	113	38	that	that	SCONJ
cril-249	113	39	we	we	PRON
cril-249	113	40	will	will	AUX
cril-249	113	41	see	see	VERB
cril-249	113	42	further	further	ADJ
cril-249	113	43	improvement	improvement	NOUN
cril-249	113	44	in	in	ADP
cril-249	113	45	the	the	DET
cril-249	113	46	performance	performance	NOUN
cril-249	113	47	of	of	ADP
cril-249	113	48	our	our	PRON
cril-249	113	49	present	present	ADJ
cril-249	113	50	day	day	NOUN
cril-249	113	51	classifier	classifier	NOUN
cril-249	113	52	by	by	ADP
cril-249	113	53	improving	improve	VERB
cril-249	113	54	the	the	DET
cril-249	113	55	quality	quality	NOUN
cril-249	113	56	of	of	ADP
cril-249	113	57	our	our	PRON
cril-249	113	58	training	training	NOUN
cril-249	113	59	data	datum	NOUN
cril-249	113	60	.	.	PUNCT
cril-249	114	1	suggestions	suggestion	NOUN
cril-249	114	2	for	for	ADP
cril-249	114	3	such	such	ADJ
cril-249	114	4	improvements	improvement	NOUN
cril-249	114	5	have	have	AUX
cril-249	114	6	included	include	VERB
cril-249	114	7	the	the	DET
cril-249	114	8	addition	addition	NOUN
cril-249	114	9	of	of	ADP
cril-249	114	10	non	non	ADJ
cril-249	114	11	-	-	ADJ
cril-249	114	12	name	name	ADJ
cril-249	114	13	based	base	VERB
cril-249	114	14	russian	russian	ADJ
cril-249	114	15	and	and	CCONJ
cril-249	114	16	arabic	arabic	ADJ
cril-249	114	17	data	datum	NOUN
cril-249	114	18	,	,	PUNCT
cril-249	114	19	comparable	comparable	ADJ
cril-249	114	20	to	to	ADP
cril-249	114	21	the	the	DET
cril-249	114	22	english	english	PROPN
cril-249	114	23	brown	brown	PROPN
cril-249	114	24	corpus	corpus	PROPN
cril-249	114	25	training	training	NOUN
cril-249	114	26	data	datum	NOUN
cril-249	114	27	being	be	AUX
cril-249	114	28	used	use	VERB
cril-249	114	29	.	.	PUNCT
cril-249	115	1	it	it	PRON
cril-249	115	2	may	may	AUX
cril-249	115	3	also	also	ADV
cril-249	115	4	be	be	AUX
cril-249	115	5	possible	possible	ADJ
cril-249	115	6	to	to	PART
cril-249	115	7	improve	improve	VERB
cril-249	115	8	performance	performance	NOUN
cril-249	115	9	by	by	ADP
cril-249	115	10	utilizing	utilize	VERB
cril-249	115	11	a	a	DET
cril-249	115	12	filtered	filter	VERB
cril-249	115	13	russian	russian	ADJ
cril-249	115	14	-	-	PUNCT
cril-249	115	15	name	name	NOUN
cril-249	115	16	data	datum	NOUN
cril-249	115	17	set	set	VERB
cril-249	115	18	,	,	PUNCT
cril-249	115	19	pared	pare	VERB
cril-249	115	20	down	down	ADP
cril-249	115	21	from	from	ADP
cril-249	115	22	the	the	DET
cril-249	115	23	large	large	ADJ
cril-249	115	24	and	and	CCONJ
cril-249	115	25	potentially	potentially	ADV
cril-249	115	26	noisy	noisy	ADJ
cril-249	115	27	set	set	NOUN
cril-249	115	28	we	we	PRON
cril-249	115	29	are	be	AUX
cril-249	115	30	currently	currently	ADV
cril-249	115	31	using	use	VERB
cril-249	115	32	.	.	PUNCT
cril-249	116	1	the	the	DET
cril-249	116	2	algorithms	algorithm	NOUN
cril-249	116	3	used	use	VERB
cril-249	116	4	in	in	ADP
cril-249	116	5	generating	generate	VERB
cril-249	116	6	the	the	DET
cril-249	116	7	letter	letter	NOUN
cril-249	116	8	-	-	PUNCT
cril-249	116	9	to	to	ADP
cril-249	116	10	-	-	PUNCT
cril-249	116	11	sound	sound	NOUN
cril-249	116	12	rules	rule	NOUN
cril-249	116	13	are	be	AUX
cril-249	116	14	those	those	PRON
cril-249	116	15	commonly	commonly	ADV
cril-249	116	16	accepted	accept	VERB
cril-249	116	17	as	as	ADP
cril-249	116	18	standards	standard	NOUN
cril-249	116	19	.	.	PUNCT
cril-249	117	1	however	however	ADV
cril-249	117	2	,	,	PUNCT
cril-249	117	3	we	we	PRON
cril-249	117	4	expect	expect	VERB
cril-249	117	5	that	that	SCONJ
cril-249	117	6	augmenting	augment	VERB
cril-249	117	7	the	the	DET
cril-249	117	8	size	size	NOUN
cril-249	117	9	of	of	ADP
cril-249	117	10	our	our	PRON
cril-249	117	11	transcribed	transcribed	ADJ
cril-249	117	12	non	non	ADJ
cril-249	117	13	-	-	ADJ
cril-249	117	14	english	english	ADJ
cril-249	117	15	data	datum	NOUN
cril-249	117	16	would	would	AUX
cril-249	117	17	increase	increase	VERB
cril-249	117	18	the	the	DET
cril-249	117	19	quality	quality	NOUN
cril-249	117	20	of	of	ADP
cril-249	117	21	output	output	NOUN
cril-249	117	22	.	.	PUNCT
cril-249	118	1	this	this	PRON
cril-249	118	2	will	will	AUX
cril-249	118	3	be	be	AUX
cril-249	118	4	one	one	NUM
cril-249	118	5	of	of	ADP
cril-249	118	6	the	the	DET
cril-249	118	7	first	first	ADJ
cril-249	118	8	areas	area	NOUN
cril-249	118	9	we	we	PRON
cril-249	118	10	will	will	AUX
cril-249	118	11	seek	seek	VERB
cril-249	118	12	to	to	PART
cril-249	118	13	retune	retune	VERB
cril-249	118	14	as	as	SCONJ
cril-249	118	15	it	it	PRON
cril-249	118	16	requires	require	VERB
cril-249	118	17	only	only	ADV
cril-249	118	18	getting	get	VERB
cril-249	118	19	access	access	NOUN
cril-249	118	20	to	to	ADP
cril-249	118	21	more	more	ADJ
cril-249	118	22	transliterations	transliteration	NOUN
cril-249	118	23	,	,	PUNCT
cril-249	118	24	transcribing	transcribe	VERB
cril-249	118	25	them	they	PRON
cril-249	118	26	into	into	ADP
cril-249	118	27	phonemes	phoneme	NOUN
cril-249	118	28	,	,	PUNCT
cril-249	118	29	and	and	CCONJ
cril-249	118	30	incorporating	incorporate	VERB
cril-249	118	31	this	this	DET
cril-249	118	32	new	new	ADJ
cril-249	118	33	data	datum	NOUN
cril-249	118	34	into	into	ADP
cril-249	118	35	our	our	PRON
cril-249	118	36	training	training	NOUN
cril-249	118	37	set	set	NOUN
cril-249	118	38	.	.	PUNCT
cril-249	119	1	finally	finally	ADV
cril-249	119	2	,	,	PUNCT
cril-249	119	3	our	our	PRON
cril-249	119	4	system	system	NOUN
cril-249	119	5	could	could	AUX
cril-249	119	6	easily	easily	ADV
cril-249	119	7	be	be	AUX
cril-249	119	8	extended	extend	VERB
cril-249	119	9	to	to	PART
cril-249	119	10	handle	handle	VERB
cril-249	119	11	other	other	ADJ
cril-249	119	12	non	non	ADJ
cril-249	119	13	-	-	ADJ
cril-249	119	14	english	english	ADJ
cril-249	119	15	word	word	NOUN
cril-249	119	16	sets	set	NOUN
cril-249	119	17	via	via	ADP
cril-249	119	18	the	the	DET
cril-249	119	19	implementation	implementation	NOUN
cril-249	119	20	of	of	ADP
cril-249	119	21	additional	additional	ADJ
cril-249	119	22	language	language	NOUN
cril-249	119	23	specific	specific	ADJ
cril-249	119	24	lts	lts	NOUN
cril-249	119	25	classifiers	classifier	NOUN
cril-249	119	26	,	,	PUNCT
cril-249	119	27	and	and	CCONJ
cril-249	119	28	by	by	ADP
cril-249	119	29	adding	add	VERB
cril-249	119	30	additional	additional	ADJ
cril-249	119	31	classification	classification	NOUN
cril-249	119	32	groups	group	NOUN
cril-249	119	33	for	for	ADP
cril-249	119	34	categorizing	categorize	VERB
cril-249	119	35	among	among	ADP
cril-249	119	36	a	a	DET
cril-249	119	37	different	different	ADJ
cril-249	119	38	set	set	NOUN
cril-249	119	39	of	of	ADP
cril-249	119	40	languages	language	NOUN
cril-249	119	41	.	.	PUNCT
cril-249	120	1	the	the	DET
cril-249	120	2	functioning	functioning	NOUN
cril-249	120	3	of	of	ADP
cril-249	120	4	our	our	PRON
cril-249	120	5	system	system	NOUN
cril-249	120	6	requires	require	VERB
cril-249	120	7	that	that	SCONJ
cril-249	120	8	there	there	PRON
cril-249	120	9	exist	exist	VERB
cril-249	120	10	a	a	DET
cril-249	120	11	degree	degree	NOUN
cril-249	120	12	of	of	ADP
cril-249	120	13	statistical	statistical	ADJ
cril-249	120	14	distinction	distinction	NOUN
cril-249	120	15	between	between	ADP
cril-249	120	16	languages	language	NOUN
cril-249	120	17	in	in	ADP
cril-249	120	18	order	order	NOUN
cril-249	120	19	to	to	PART
cril-249	120	20	effectively	effectively	ADV
cril-249	120	21	differentiate	differentiate	VERB
cril-249	120	22	between	between	ADP
cril-249	120	23	them	they	PRON
cril-249	120	24	.	.	PUNCT
cril-249	121	1	future	future	ADJ
cril-249	121	2	work	work	NOUN
cril-249	121	3	could	could	AUX
cril-249	121	4	easily	easily	ADV
cril-249	121	5	address	address	VERB
cril-249	121	6	the	the	DET
cril-249	121	7	theoretical	theoretical	ADJ
cril-249	121	8	question	question	NOUN
cril-249	121	9	of	of	ADP
cril-249	121	10	assessing	assess	VERB
cril-249	121	11	the	the	DET
cril-249	121	12	limits	limit	NOUN
cril-249	121	13	on	on	ADP
cril-249	121	14	the	the	DET
cril-249	121	15	number	number	NOUN
cril-249	121	16	of	of	ADP
cril-249	121	17	languages	language	NOUN
cril-249	121	18	it	it	PRON
cril-249	121	19	can	can	AUX
cril-249	121	20	handle	handle	AUX
cril-249	121	21	based	base	VERB
cril-249	121	22	upon	upon	SCONJ
cril-249	121	23	their	their	PRON
cril-249	121	24	phonemic	phonemic	ADJ
cril-249	121	25	similarity	similarity	NOUN
cril-249	121	26	and	and	CCONJ
cril-249	121	27	difference	difference	NOUN
cril-249	121	28	.	.	PUNCT
cril-249	122	1	additionally	additionally	ADV
cril-249	122	2	this	this	PRON
cril-249	122	3	would	would	AUX
cril-249	122	4	serve	serve	VERB
cril-249	122	5	to	to	PART
cril-249	122	6	test	test	VERB
cril-249	122	7	whether	whether	SCONJ
cril-249	122	8	the	the	DET
cril-249	122	9	system	system	NOUN
cril-249	122	10	is	be	AUX
cril-249	122	11	robust	robust	ADJ
cril-249	122	12	enough	enough	ADV
cril-249	122	13	to	to	PART
cril-249	122	14	distinguish	distinguish	VERB
cril-249	122	15	between	between	ADP
cril-249	122	16	more	more	ADV
cril-249	122	17	closely	closely	ADV
cril-249	122	18	related	relate	VERB
cril-249	122	19	languages	language	NOUN
cril-249	122	20	.	.	PUNCT
cril-249	123	1	references	reference	NOUN
cril-249	123	2	coker	coker	NOUN
cril-249	123	3	,	,	PUNCT
cril-249	123	4	k.	k.	PROPN
cril-249	123	5	church	church	PROPN
cril-249	123	6	and	and	CCONJ
cril-249	123	7	m.	m.	NOUN
cril-249	123	8	liberman	liberman	PROPN
cril-249	123	9	.	.	PUNCT
cril-249	124	1	1990	1990	NUM
cril-249	124	2	.	.	PUNCT
cril-249	125	1	'	'	PUNCT
cril-249	125	2	morphology	morphology	NOUN
cril-249	125	3	and	and	CCONJ
cril-249	125	4	rhyming	rhyme	VERB
cril-249	125	5	:	:	PUNCT
cril-249	125	6	two	two	NUM
cril-249	125	7	powerful	powerful	ADJ
cril-249	125	8	alternatives	alternative	NOUN
cril-249	125	9	to	to	ADP
cril-249	125	10	letter	letter	NOUN
cril-249	125	11	-	-	PUNCT
cril-249	125	12	to	to	ADP
cril-249	125	13	-	-	PUNCT
cril-249	125	14	sound	sound	NOUN
cril-249	125	15	rules	rule	NOUN
cril-249	125	16	for	for	ADP
cril-249	125	17	speech	speech	NOUN
cril-249	125	18	synthesis	synthesis	NOUN
cril-249	125	19	.	.	PUNCT
cril-249	125	20	'	'	PUNCT
cril-249	126	1	european	european	ADJ
cril-249	126	2	speech	speech	NOUN
cril-249	126	3	communication	communication	NOUN
cril-249	126	4	association	association	NOUN
cril-249	126	5	,	,	PUNCT
cril-249	126	6	conference	conference	NOUN
cril-249	126	7	on	on	ADP
cril-249	126	8	speech	speech	NOUN
cril-249	126	9	synthesis	synthesis	NOUN
cril-249	126	10	.	.	PUNCT
cril-249	127	1	jurafsky	jurafsky	PROPN
cril-249	127	2	,	,	PUNCT
cril-249	127	3	daniel	daniel	PROPN
cril-249	127	4	and	and	CCONJ
cril-249	127	5	jim	jim	PROPN
cril-249	127	6	martin	martin	PROPN
cril-249	127	7	.	.	PROPN
cril-249	128	1	2000	2000	NUM
cril-249	128	2	.	.	PUNCT
cril-249	129	1	speech	speech	NOUN
cril-249	129	2	and	and	CCONJ
cril-249	129	3	language	language	NOUN
cril-249	129	4	processing	processing	NOUN
cril-249	129	5	.	.	PUNCT
cril-249	130	1	prentice	prentice	PROPN
cril-249	130	2	hall	hall	PROPN
cril-249	130	3	.	.	PUNCT
cril-249	131	1	pellom	pellom	NOUN
cril-249	131	2	,	,	PUNCT
cril-249	131	3	bryan	bryan	PROPN
cril-249	131	4	and	and	CCONJ
cril-249	131	5	kadri	kadri	PROPN
cril-249	131	6	hacioglu	hacioglu	PROPN
cril-249	131	7	.	.	PROPN
cril-249	131	8	2003	2003	NUM
cril-249	131	9	.	.	PUNCT
cril-249	132	1	sonic	sonic	ADJ
cril-249	132	2	:	:	PUNCT
cril-249	132	3	the	the	DET
cril-249	132	4	university	university	PROPN
cril-249	132	5	of	of	ADP
cril-249	132	6	colorado	colorado	PROPN
cril-249	132	7	continuous	continuous	ADJ
cril-249	132	8	speech	speech	NOUN
cril-249	132	9	recognizer	recognizer	NOUN
cril-249	132	10	,	,	PUNCT
cril-249	132	11	technical	technical	ADJ
cril-249	132	12	report	report	NOUN
cril-249	132	13	tr	tr	VERB
cril-249	132	14	-	-	ADJ
cril-249	132	15	cslr-2001	cslr-2001	ADJ
cril-249	132	16	-	-	PUNCT
cril-249	132	17	01	01	NUM
cril-249	132	18	.	.	PUNCT
cril-249	132	19	reynar	reynar	PROPN
cril-249	132	20	,	,	PUNCT
cril-249	132	21	jeffrey	jeffrey	PROPN
cril-249	132	22	c.	c.	PROPN
cril-249	132	23	and	and	CCONJ
cril-249	132	24	adwait	adwait	PROPN
cril-249	132	25	ratnaparkhi	ratnaparkhi	PROPN
cril-249	132	26	.	.	PUNCT
cril-249	133	1	1997	1997	NUM
cril-249	133	2	.	.	PUNCT
cril-249	134	1	'	'	PUNCT
cril-249	134	2	a	a	DET
cril-249	134	3	maximum	maximum	PROPN
cril-249	134	4	entropy	entropy	NOUN
cril-249	134	5	approach	approach	NOUN
cril-249	134	6	to	to	ADP
cril-249	134	7	identifying	identify	VERB
cril-249	134	8	sentence	sentence	NOUN
cril-249	134	9	boundaries	boundary	NOUN
cril-249	134	10	.	.	PUNCT
cril-249	134	11	'	'	PUNCT
cril-249	135	1	in	in	ADP
cril-249	135	2	proceedings	proceeding	NOUN
cril-249	135	3	of	of	ADP
cril-249	135	4	the	the	DET
cril-249	135	5	fifth	fifth	ADJ
cril-249	135	6	conference	conference	NOUN
cril-249	135	7	on	on	ADP
cril-249	135	8	applied	apply	VERB
cril-249	135	9	natural	natural	ADJ
cril-249	135	10	language	language	NOUN
cril-249	135	11	processing	processing	NOUN
cril-249	135	12	.	.	PUNCT
cril-249	136	1	russell	russell	PROPN
cril-249	136	2	,	,	PUNCT
cril-249	136	3	stuart	stuart	PROPN
cril-249	136	4	and	and	CCONJ
cril-249	136	5	peter	peter	PROPN
cril-249	136	6	norvig	norvig	PROPN
cril-249	136	7	.	.	PUNCT
cril-249	137	1	2002	2002	NUM
cril-249	137	2	.	.	PUNCT
cril-249	138	1	artificial	artificial	ADJ
cril-249	138	2	intelligence	intelligence	NOUN
cril-249	138	3	:	:	PUNCT
cril-249	138	4	a	a	DET
cril-249	138	5	modern	modern	ADJ
cril-249	138	6	approach	approach	NOUN
cril-249	138	7	.	.	PUNCT
cril-249	139	1	prentice	prentice	PROPN
cril-249	139	2	hall	hall	PROPN
cril-249	139	3	.	.	PUNCT
cril-249	140	1	urls	url	NOUN
cril-249	140	2	goldschmidt	goldschmidt	PROPN
cril-249	140	3	,	,	PUNCT
cril-249	140	4	paul	paul	PROPN
cril-249	140	5	.	.	PROPN
cril-249	140	6	1996	1996	NUM
cril-249	140	7	.	.	PUNCT
cril-249	141	1	a	a	DET
cril-249	141	2	dictionary	dictionary	NOUN
cril-249	141	3	of	of	ADP
cril-249	141	4	period	period	NOUN
cril-249	141	5	russian	russian	ADJ
cril-249	141	6	names	name	NOUN
cril-249	141	7	.	.	PUNCT
cril-249	142	1	http://www.sca.org/heraldry/paul/.	http://www.sca.org/heraldry/paul/.	PROPN
cril-249	142	2	retrieved	retrieve	VERB
cril-249	142	3	december	december	PROPN
cril-249	142	4	2003	2003	NUM
cril-249	142	5	.	.	PUNCT
cril-249	143	1	cmu	cmu	PROPN
cril-249	143	2	sphinx	sphinx	PROPN
cril-249	143	3	.	.	PUNCT
cril-249	144	1	http://fife.speech.cs.cmu.edu/sphinx/.	http://fife.speech.cs.cmu.edu/sphinx/.	X
cril-249	144	2	retrieved	retrieve	VERB
cril-249	144	3	december	december	PROPN
cril-249	144	4	2003	2003	NUM
cril-249	144	5	.	.	PUNCT
cril-249	145	1	8	8	NUM
cril-249	145	2	colorado	colorado	NOUN
cril-249	145	3	research	research	NOUN
cril-249	145	4	in	in	ADP
cril-249	145	5	linguistics	linguistic	NOUN
cril-249	145	6	,	,	PUNCT
cril-249	145	7	vol	vol	NOUN
cril-249	145	8	.	.	PROPN
cril-249	145	9	17	17	NUM
cril-249	146	1	[	[	X
cril-249	146	2	2004	2004	NUM
cril-249	146	3	]	]	PUNCT
cril-249	146	4	https://scholar.colorado.edu/cril/vol17/iss1/6	https://scholar.colorado.edu/cril/vol17/iss1/6	NOUN
cril-249	146	5	doi	doi	NOUN
cril-249	146	6	:	:	PUNCT
cril-249	146	7	https://doi.org/10.25810/60mf-fn94	https://doi.org/10.25810/60mf-fn94	PROPN
cril-249	146	8	http://www.sca.org/heraldry/paul/	http://www.sca.org/heraldry/paul/	PROPN
cril-249	146	9	http://fife.speech.cs.cmu.edu/sphinx/	http://fife.speech.cs.cmu.edu/sphinx/	NUM
cril-249	146	10	colorado	colorado	NOUN
cril-249	146	11	research	research	NOUN
cril-249	146	12	in	in	ADP
cril-249	146	13	linguistics	linguistic	NOUN
cril-249	146	14	6	6	NUM
cril-249	146	15	-	-	SYM
cril-249	146	16	2004	2004	NUM
cril-249	146	17	language	language	NOUN
cril-249	146	18	identification	identification	NOUN
cril-249	146	19	and	and	CCONJ
cril-249	146	20	language	language	NOUN
cril-249	146	21	specific	specific	ADJ
cril-249	146	22	letter	letter	NOUN
cril-249	146	23	-	-	PUNCT
cril-249	146	24	to	to	ADP
cril-249	146	25	-	-	PUNCT
cril-249	146	26	sound	sound	NOUN
cril-249	146	27	rules	rule	NOUN
cril-249	146	28	stephen	stephen	PROPN
cril-249	146	29	lewis	lewis	PROPN
cril-249	146	30	katie	katie	PROPN
cril-249	146	31	mcgrath	mcgrath	PROPN
cril-249	146	32	jeffrey	jeffrey	PROPN
cril-249	146	33	reuppel	reuppel	PROPN
cril-249	146	34	recommended	recommend	VERB
cril-249	146	35	citation	citation	NOUN
cril-249	146	36	introduction	introduction	NOUN
cril-249	146	37	past	past	ADP
cril-249	146	38	/	/	SYM
cril-249	146	39	related	relate	VERB
cril-249	146	40	work	work	NOUN
cril-249	146	41	language	language	NOUN
cril-249	146	42	classification	classification	NOUN
cril-249	146	43	training	training	NOUN
cril-249	146	44	data	data	NOUN
cril-249	146	45	classification	classification	NOUN
cril-249	146	46	algorithm	algorithm	NOUN
cril-249	146	47	letter	letter	NOUN
cril-249	146	48	to	to	PART
cril-249	146	49	sound	sound	VERB
cril-249	146	50	rules	rule	NOUN
cril-249	146	51	training	train	VERB
cril-249	146	52	data	datum	NOUN
cril-249	146	53	training	training	NOUN
cril-249	146	54	algorithm	algorithm	NOUN
cril-249	146	55	results	result	VERB
cril-249	146	56	/	/	SYM
cril-249	146	57	analysis	analysis	NOUN
cril-249	146	58	phase	phase	NOUN
cril-249	146	59	1	1	NUM
cril-249	146	60	:	:	PUNCT
cril-249	146	61	analysis	analysis	NOUN
cril-249	146	62	of	of	ADP
cril-249	146	63	the	the	DET
cril-249	146	64	classifier	classifier	NOUN
cril-249	146	65	phase	phase	NOUN
cril-249	146	66	2	2	NUM
cril-249	146	67	:	:	PUNCT
cril-249	146	68	analysis	analysis	NOUN
cril-249	146	69	of	of	ADP
cril-249	146	70	the	the	DET
cril-249	146	71	letter	letter	NOUN
cril-249	146	72	-	-	PUNCT
cril-249	146	73	to	to	ADP
cril-249	146	74	-	-	PUNCT
cril-249	146	75	sound	sound	NOUN
cril-249	146	76	rules	rule	NOUN
cril-249	146	77	phase	phase	NOUN
cril-249	146	78	3	3	NUM
cril-249	146	79	:	:	PUNCT
cril-249	146	80	analysis	analysis	NOUN
cril-249	146	81	of	of	ADP
cril-249	146	82	the	the	DET
cril-249	146	83	complete	complete	ADJ
cril-249	146	84	system	system	NOUN
cril-249	146	85	future	future	ADJ
cril-249	146	86	work	work	NOUN
cril-249	146	87	and	and	CCONJ
cril-249	146	88	conclusions	conclusion	NOUN