id	sid	tid	token	lemma	pos
bracis-33560	1	1	aroeira	aroeira	PROPN
bracis-33560	1	2	:	:	PUNCT
bracis-33560	1	3	a	a	DET
bracis-33560	1	4	curated	curate	VERB
bracis-33560	1	5	corpus	corpus	NOUN
bracis-33560	1	6	for	for	ADP
bracis-33560	1	7	 	 	SPACE
bracis-33560	1	8	the	the	DET
bracis-33560	1	9	 	 	SPACE
bracis-33560	1	10	portuguese	portuguese	ADJ
bracis-33560	1	11	language	language	NOUN
bracis-33560	1	12	with	with	ADP
bracis-33560	1	13	 	 	SPACE
bracis-33560	1	14	a	a	DET
bracis-33560	1	15	 	 	SPACE
bracis-33560	1	16	large	large	ADJ
bracis-33560	1	17	number	number	NOUN
bracis-33560	1	18	of	of	ADP
bracis-33560	1	19	 	 	SPACE
bracis-33560	1	20	tokens	token	NOUN
bracis-33560	1	21	|	|	ADV
bracis-33560	1	22	springer	springer	NOUN
bracis-33560	1	23	nature	nature	PROPN
bracis-33560	1	24	link	link	PROPN
bracis-33560	1	25	(	(	PUNCT
bracis-33560	1	26	formerly	formerly	ADV
bracis-33560	1	27	springerlink	springerlink	NOUN
bracis-33560	1	28	)	)	PUNCT
bracis-33560	1	29	skip	skip	VERB
bracis-33560	1	30	to	to	ADP
bracis-33560	1	31	main	main	ADJ
bracis-33560	1	32	content	content	NOUN
bracis-33560	1	33	advertisement	advertisement	NOUN
bracis-33560	1	34	log	log	NOUN
bracis-33560	1	35	in	in	ADP
bracis-33560	1	36	menu	menu	NOUN
bracis-33560	1	37	find	find	VERB
bracis-33560	1	38	a	a	DET
bracis-33560	1	39	journal	journal	NOUN
bracis-33560	1	40	publish	publish	VERB
bracis-33560	1	41	with	with	ADP
bracis-33560	1	42	us	we	PRON
bracis-33560	1	43	track	track	VERB
bracis-33560	1	44	your	your	PRON
bracis-33560	1	45	research	research	NOUN
bracis-33560	1	46	search	search	NOUN
bracis-33560	1	47	cart	cart	NOUN
bracis-33560	1	48	home	home	NOUN
bracis-33560	1	49	intelligent	intelligent	ADJ
bracis-33560	1	50	systems	system	NOUN
bracis-33560	1	51	conference	conference	NOUN
bracis-33560	1	52	paper	paper	NOUN
bracis-33560	1	53	aroeira	aroeira	PROPN
bracis-33560	1	54	:	:	PUNCT
bracis-33560	1	55	a	a	DET
bracis-33560	1	56	curated	curate	VERB
bracis-33560	1	57	corpus	corpus	NOUN
bracis-33560	1	58	for	for	ADP
bracis-33560	1	59	 	 	SPACE
bracis-33560	1	60	the	the	DET
bracis-33560	1	61	 	 	SPACE
bracis-33560	1	62	portuguese	portuguese	ADJ
bracis-33560	1	63	language	language	NOUN
bracis-33560	1	64	with	with	ADP
bracis-33560	1	65	 	 	SPACE
bracis-33560	1	66	a	a	DET
bracis-33560	1	67	 	 	SPACE
bracis-33560	1	68	large	large	ADJ
bracis-33560	1	69	number	number	NOUN
bracis-33560	1	70	of	of	ADP
bracis-33560	1	71	 	 	SPACE
bracis-33560	1	72	tokens	token	NOUN
bracis-33560	1	73	conference	conference	NOUN
bracis-33560	1	74	paper	paper	NOUN
bracis-33560	1	75	first	first	ADV
bracis-33560	1	76	online	online	ADV
bracis-33560	1	77	:	:	PUNCT
bracis-33560	1	78	30	30	NUM
bracis-33560	1	79	january	january	NOUN
bracis-33560	1	80	2025	2025	NUM
bracis-33560	1	81	pp	pp	ADP
bracis-33560	1	82	185–199	185–199	NUM
bracis-33560	1	83	cite	cite	VERB
bracis-33560	1	84	this	this	DET
bracis-33560	1	85	conference	conference	NOUN
bracis-33560	1	86	paper	paper	NOUN
bracis-33560	1	87	access	access	NOUN
bracis-33560	1	88	provided	provide	VERB
bracis-33560	1	89	by	by	ADP
bracis-33560	1	90	university	university	PROPN
bracis-33560	1	91	of	of	ADP
bracis-33560	1	92	notre	notre	PROPN
bracis-33560	1	93	dame	dame	PROPN
bracis-33560	1	94	hesburgh	hesburgh	PROPN
bracis-33560	1	95	library	library	PROPN
bracis-33560	1	96	download	download	PROPN
bracis-33560	1	97	book	book	NOUN
bracis-33560	1	98	pdf	pdf	PROPN
bracis-33560	1	99	download	download	NOUN
bracis-33560	1	100	book	book	NOUN
bracis-33560	1	101	epub	epub	PROPN
bracis-33560	1	102	intelligent	intelligent	ADJ
bracis-33560	1	103	systems	system	NOUN
bracis-33560	1	104	(	(	PUNCT
bracis-33560	1	105	bracis	bracis	NOUN
bracis-33560	1	106	2024	2024	NUM
bracis-33560	1	107	)	)	PUNCT
bracis-33560	1	108	aroeira	aroeira	PROPN
bracis-33560	1	109	:	:	PUNCT
bracis-33560	1	110	a	a	DET
bracis-33560	1	111	curated	curate	VERB
bracis-33560	1	112	corpus	corpus	NOUN
bracis-33560	1	113	for	for	ADP
bracis-33560	1	114	 	 	SPACE
bracis-33560	1	115	the	the	DET
bracis-33560	1	116	 	 	SPACE
bracis-33560	1	117	portuguese	portuguese	ADJ
bracis-33560	1	118	language	language	NOUN
bracis-33560	1	119	with	with	ADP
bracis-33560	1	120	 	 	SPACE
bracis-33560	1	121	a	a	DET
bracis-33560	1	122	 	 	SPACE
bracis-33560	1	123	large	large	ADJ
bracis-33560	1	124	number	number	NOUN
bracis-33560	1	125	of	of	ADP
bracis-33560	1	126	 	 	SPACE
bracis-33560	1	127	tokens	token	NOUN
bracis-33560	1	128	download	download	VERB
bracis-33560	1	129	book	book	NOUN
bracis-33560	1	130	pdf	pdf	PROPN
bracis-33560	1	131	download	download	NOUN
bracis-33560	1	132	book	book	PROPN
bracis-33560	1	133	epub	epub	PROPN
bracis-33560	1	134	thiago	thiago	PROPN
bracis-33560	1	135	lira9	lira9	PROPN
bracis-33560	1	136	,	,	PUNCT
bracis-33560	1	137	flávio	flávio	NOUN
bracis-33560	1	138	cação9	cação9	PROPN
bracis-33560	1	139	,	,	PUNCT
bracis-33560	1	140	cinthia	cinthia	NOUN
bracis-33560	1	141	souza9	souza9	PROPN
bracis-33560	1	142	,	,	PUNCT
bracis-33560	1	143	joão	joão	PROPN
bracis-33560	1	144	valentini9	valentini9	PROPN
bracis-33560	1	145	,	,	PUNCT
bracis-33560	1	146	edson	edson	PROPN
bracis-33560	1	147	bollis9	bollis9	PROPN
bracis-33560	1	148	,	,	PUNCT
bracis-33560	1	149	otavio	otavio	NOUN
bracis-33560	1	150	oliveira9	oliveira9	NOUN
bracis-33560	1	151	,	,	PUNCT
bracis-33560	1	152	renato	renato	PROPN
bracis-33560	1	153	almeida9	almeida9	PROPN
bracis-33560	1	154	,	,	PUNCT
bracis-33560	1	155	marcio	marcio	PROPN
bracis-33560	1	156	magalhães9	magalhães9	PROPN
bracis-33560	1	157	,	,	PUNCT
bracis-33560	1	158	katia	katia	PROPN
bracis-33560	1	159	poloni9	poloni9	PROPN
bracis-33560	1	160	,	,	PUNCT
bracis-33560	1	161	andre	andre	PROPN
bracis-33560	1	162	oliveira9	oliveira9	PROPN
bracis-33560	1	163	&	&	CCONJ
bracis-33560	1	164	…	…	PUNCT
bracis-33560	1	165	lucas	lucas	PROPN
bracis-33560	1	166	pellicer9	pellicer9	NOUN
bracis-33560	1	167	  	  	SPACE
bracis-33560	1	168	show	show	NOUN
bracis-33560	1	169	authors	author	NOUN
bracis-33560	1	170	part	part	NOUN
bracis-33560	1	171	of	of	ADP
bracis-33560	1	172	the	the	DET
bracis-33560	1	173	book	book	NOUN
bracis-33560	1	174	series	series	NOUN
bracis-33560	1	175	:	:	PUNCT
bracis-33560	1	176	lecture	lecture	NOUN
bracis-33560	1	177	notes	note	NOUN
bracis-33560	1	178	in	in	ADP
bracis-33560	1	179	computer	computer	NOUN
bracis-33560	1	180	science	science	NOUN
bracis-33560	1	181	(	(	PUNCT
bracis-33560	1	182	(	(	PUNCT
bracis-33560	1	183	lnai	lnai	ADJ
bracis-33560	1	184	,	,	PUNCT
bracis-33560	1	185	volume	volume	NOUN
bracis-33560	1	186	15412	15412	NUM
bracis-33560	1	187	)	)	PUNCT
bracis-33560	1	188	)	)	PUNCT
bracis-33560	1	189	included	include	VERB
bracis-33560	1	190	in	in	ADP
bracis-33560	1	191	the	the	DET
bracis-33560	1	192	following	follow	VERB
bracis-33560	1	193	conference	conference	NOUN
bracis-33560	1	194	series	series	NOUN
bracis-33560	1	195	:	:	PUNCT
bracis-33560	1	196	brazilian	brazilian	ADJ
bracis-33560	1	197	conference	conference	NOUN
bracis-33560	1	198	on	on	ADP
bracis-33560	1	199	intelligent	intelligent	ADJ
bracis-33560	1	200	systems	system	NOUN
bracis-33560	1	201	554	554	NUM
bracis-33560	1	202	accesses	access	VERB
bracis-33560	1	203	1	1	NUM
bracis-33560	1	204	altmetric	altmetric	ADJ
bracis-33560	1	205	abstract	abstract	NOUN
bracis-33560	1	206	the	the	DET
bracis-33560	1	207	emphasis	emphasis	NOUN
bracis-33560	1	208	on	on	ADP
bracis-33560	1	209	constructing	construct	VERB
bracis-33560	1	210	extensive	extensive	ADJ
bracis-33560	1	211	datasets	dataset	NOUN
bracis-33560	1	212	for	for	ADP
bracis-33560	1	213	training	train	VERB
bracis-33560	1	214	large	large	ADJ
bracis-33560	1	215	language	language	NOUN
bracis-33560	1	216	models	model	NOUN
bracis-33560	1	217	(	(	PUNCT
bracis-33560	1	218	llm	llm	NOUN
bracis-33560	1	219	)	)	PUNCT
bracis-33560	1	220	has	have	AUX
bracis-33560	1	221	recently	recently	ADV
bracis-33560	1	222	increased	increase	VERB
bracis-33560	1	223	,	,	PUNCT
bracis-33560	1	224	and	and	CCONJ
bracis-33560	1	225	current	current	ADJ
bracis-33560	1	226	literature	literature	NOUN
bracis-33560	1	227	predominantly	predominantly	ADV
bracis-33560	1	228	features	feature	VERB
bracis-33560	1	229	datasets	dataset	NOUN
bracis-33560	1	230	for	for	ADP
bracis-33560	1	231	high	high	ADJ
bracis-33560	1	232	-	-	PUNCT
bracis-33560	1	233	resource	resource	NOUN
bracis-33560	1	234	languages	language	NOUN
bracis-33560	1	235	such	such	ADJ
bracis-33560	1	236	as	as	ADP
bracis-33560	1	237	english	english	PROPN
bracis-33560	1	238	and	and	CCONJ
bracis-33560	1	239	chinese	chinese	PROPN
bracis-33560	1	240	.	.	PUNCT
bracis-33560	2	1	however	however	ADV
bracis-33560	2	2	,	,	PUNCT
bracis-33560	2	3	there	there	PRON
bracis-33560	2	4	is	be	VERB
bracis-33560	2	5	a	a	DET
bracis-33560	2	6	notable	notable	ADJ
bracis-33560	2	7	scarcity	scarcity	NOUN
bracis-33560	2	8	of	of	ADP
bracis-33560	2	9	high	high	ADJ
bracis-33560	2	10	-	-	PUNCT
bracis-33560	2	11	quality	quality	NOUN
bracis-33560	2	12	corpora	corpora	NOUN
bracis-33560	2	13	for	for	ADP
bracis-33560	2	14	the	the	DET
bracis-33560	2	15	portuguese	portuguese	ADJ
bracis-33560	2	16	language	language	NOUN
bracis-33560	2	17	.	.	PUNCT
bracis-33560	3	1	to	to	PART
bracis-33560	3	2	address	address	VERB
bracis-33560	3	3	this	this	DET
bracis-33560	3	4	limitation	limitation	NOUN
bracis-33560	3	5	,	,	PUNCT
bracis-33560	3	6	we	we	PRON
bracis-33560	3	7	propose	propose	VERB
bracis-33560	3	8	aroeira	aroeira	PROPN
bracis-33560	3	9	,	,	PUNCT
bracis-33560	3	10	a	a	DET
bracis-33560	3	11	curated	curate	VERB
bracis-33560	3	12	corpus	corpus	NOUN
bracis-33560	3	13	explicitly	explicitly	ADV
bracis-33560	3	14	designed	design	VERB
bracis-33560	3	15	for	for	ADP
bracis-33560	3	16	training	train	VERB
bracis-33560	3	17	large	large	ADJ
bracis-33560	3	18	language	language	NOUN
bracis-33560	3	19	models	model	NOUN
bracis-33560	3	20	in	in	ADP
bracis-33560	3	21	the	the	DET
bracis-33560	3	22	portuguese	portuguese	ADJ
bracis-33560	3	23	language	language	NOUN
bracis-33560	3	24	,	,	PUNCT
bracis-33560	3	25	with	with	ADP
bracis-33560	3	26	a	a	DET
bracis-33560	3	27	focus	focus	NOUN
bracis-33560	3	28	on	on	ADP
bracis-33560	3	29	the	the	DET
bracis-33560	3	30	brazilian	brazilian	ADJ
bracis-33560	3	31	portuguese	portuguese	PROPN
bracis-33560	3	32	one	one	NUM
bracis-33560	3	33	.	.	PUNCT
bracis-33560	4	1	the	the	DET
bracis-33560	4	2	aroeira	aroeira	PROPN
bracis-33560	4	3	corpus	corpus	PROPN
bracis-33560	4	4	consists	consist	VERB
bracis-33560	4	5	of	of	ADP
bracis-33560	4	6	100	100	NUM
bracis-33560	4	7	gb	gb	NOUN
bracis-33560	4	8	of	of	ADP
bracis-33560	4	9	texts	text	NOUN
bracis-33560	4	10	from	from	ADP
bracis-33560	4	11	various	various	ADJ
bracis-33560	4	12	internet	internet	NOUN
bracis-33560	4	13	platforms	platform	NOUN
bracis-33560	4	14	,	,	PUNCT
bracis-33560	4	15	processed	process	VERB
bracis-33560	4	16	through	through	ADP
bracis-33560	4	17	a	a	DET
bracis-33560	4	18	comprehensive	comprehensive	ADJ
bracis-33560	4	19	pipeline	pipeline	NOUN
bracis-33560	4	20	to	to	PART
bracis-33560	4	21	ensure	ensure	VERB
bracis-33560	4	22	superior	superior	ADJ
bracis-33560	4	23	quality	quality	NOUN
bracis-33560	4	24	.	.	PUNCT
bracis-33560	5	1	the	the	DET
bracis-33560	5	2	pipeline	pipeline	NOUN
bracis-33560	5	3	handles	handle	VERB
bracis-33560	5	4	downloading	download	VERB
bracis-33560	5	5	,	,	PUNCT
bracis-33560	5	6	text	text	NOUN
bracis-33560	5	7	extraction	extraction	NOUN
bracis-33560	5	8	,	,	PUNCT
bracis-33560	5	9	language	language	NOUN
bracis-33560	5	10	identification	identification	NOUN
bracis-33560	5	11	,	,	PUNCT
bracis-33560	5	12	application	application	NOUN
bracis-33560	5	13	of	of	ADP
bracis-33560	5	14	quality	quality	NOUN
bracis-33560	5	15	and	and	CCONJ
bracis-33560	5	16	bias	bias	NOUN
bracis-33560	5	17	filters	filter	NOUN
bracis-33560	5	18	,	,	PUNCT
bracis-33560	5	19	and	and	CCONJ
bracis-33560	5	20	storage	storage	NOUN
bracis-33560	5	21	,	,	PUNCT
bracis-33560	5	22	all	all	PRON
bracis-33560	5	23	tailored	tailor	VERB
bracis-33560	5	24	for	for	ADP
bracis-33560	5	25	the	the	DET
bracis-33560	5	26	portuguese	portuguese	ADJ
bracis-33560	5	27	language	language	NOUN
bracis-33560	5	28	.	.	PUNCT
bracis-33560	6	1	the	the	DET
bracis-33560	6	2	resulting	result	VERB
bracis-33560	6	3	corpus	corpus	NOUN
bracis-33560	6	4	contains	contain	VERB
bracis-33560	6	5	35.3	35.3	NUM
bracis-33560	6	6	million	million	NUM
bracis-33560	6	7	documents	document	NOUN
bracis-33560	6	8	and	and	CCONJ
bracis-33560	6	9	over	over	ADP
bracis-33560	6	10	15.1	15.1	NUM
bracis-33560	6	11	billion	billion	NUM
bracis-33560	6	12	tokens	token	NOUN
bracis-33560	6	13	,	,	PUNCT
bracis-33560	6	14	surpassing	surpass	VERB
bracis-33560	6	15	the	the	DET
bracis-33560	6	16	largest	large	ADJ
bracis-33560	6	17	previously	previously	ADV
bracis-33560	6	18	available	available	ADJ
bracis-33560	6	19	corpus	corpus	NOUN
bracis-33560	6	20	in	in	ADP
bracis-33560	6	21	this	this	DET
bracis-33560	6	22	domain	domain	NOUN
bracis-33560	6	23	.	.	PUNCT
bracis-33560	7	1	access	access	NOUN
bracis-33560	7	2	provided	provide	VERB
bracis-33560	7	3	by	by	ADP
bracis-33560	7	4	university	university	PROPN
bracis-33560	7	5	of	of	ADP
bracis-33560	7	6	notre	notre	PROPN
bracis-33560	7	7	dame	dame	PROPN
bracis-33560	7	8	hesburgh	hesburgh	PROPN
bracis-33560	7	9	library	library	PROPN
bracis-33560	7	10	.	.	PUNCT
bracis-33560	8	1	download	download	PROPN
bracis-33560	8	2	conference	conference	NOUN
bracis-33560	8	3	paper	paper	NOUN
bracis-33560	8	4	pdf	pdf	NOUN
bracis-33560	8	5	similar	similar	ADJ
bracis-33560	8	6	content	content	NOUN
bracis-33560	8	7	being	be	AUX
bracis-33560	8	8	viewed	view	VERB
bracis-33560	8	9	by	by	ADP
bracis-33560	8	10	others	other	NOUN
bracis-33560	8	11	sabiá	sabiá	PROPN
bracis-33560	8	12	:	:	PUNCT
bracis-33560	8	13	portuguese	portuguese	ADJ
bracis-33560	8	14	large	large	ADJ
bracis-33560	8	15	language	language	NOUN
bracis-33560	8	16	models	model	NOUN
bracis-33560	8	17	chapter	chapter	NOUN
bracis-33560	8	18	©	©	ADP
bracis-33560	8	19	2023	2023	NUM
bracis-33560	8	20	historical	historical	ADJ
bracis-33560	8	21	portuguese	portuguese	NOUN
bracis-33560	8	22	corpora	corpora	PROPN
bracis-33560	8	23	:	:	PUNCT
bracis-33560	8	24	a	a	DET
bracis-33560	8	25	survey	survey	NOUN
bracis-33560	8	26	article	article	NOUN
bracis-33560	8	27	open	open	ADJ
bracis-33560	8	28	access	access	NOUN
bracis-33560	8	29	18	18	NUM
bracis-33560	8	30	july	july	PROPN
bracis-33560	8	31	2024	2024	NUM
bracis-33560	8	32	mapping	mapping	NOUN
bracis-33560	8	33	languages	language	NOUN
bracis-33560	8	34	:	:	PUNCT
bracis-33560	8	35	the	the	DET
bracis-33560	8	36	corpus	corpus	NOUN
bracis-33560	8	37	of	of	ADP
bracis-33560	8	38	global	global	ADJ
bracis-33560	8	39	language	language	NOUN
bracis-33560	8	40	use	use	NOUN
bracis-33560	8	41	article	article	NOUN
bracis-33560	8	42	08	08	NUM
bracis-33560	8	43	april	april	PROPN
bracis-33560	8	44	2020	2020	NUM
bracis-33560	8	45	1	1	NUM
bracis-33560	8	46	introduction	introduction	NOUN
bracis-33560	8	47	most	most	ADJ
bracis-33560	8	48	modern	modern	ADJ
bracis-33560	8	49	natural	natural	ADJ
bracis-33560	8	50	language	language	NOUN
bracis-33560	8	51	processing	processing	NOUN
bracis-33560	8	52	(	(	PUNCT
bracis-33560	8	53	nlp	nlp	NOUN
bracis-33560	8	54	)	)	PUNCT
bracis-33560	8	55	methodologies	methodology	NOUN
bracis-33560	8	56	,	,	PUNCT
bracis-33560	8	57	including	include	VERB
bracis-33560	8	58	large	large	ADJ
bracis-33560	8	59	language	language	NOUN
bracis-33560	8	60	models	model	NOUN
bracis-33560	8	61	(	(	PUNCT
bracis-33560	8	62	llms	llm	NOUN
bracis-33560	8	63	)	)	PUNCT
bracis-33560	8	64	,	,	PUNCT
bracis-33560	8	65	rely	rely	VERB
bracis-33560	8	66	on	on	ADP
bracis-33560	8	67	extensive	extensive	ADJ
bracis-33560	8	68	text	text	NOUN
bracis-33560	8	69	corpora	corpora	NOUN
bracis-33560	8	70	for	for	ADP
bracis-33560	8	71	precise	precise	ADJ
bracis-33560	8	72	training	training	NOUN
bracis-33560	8	73	and	and	CCONJ
bracis-33560	8	74	weight	weight	NOUN
bracis-33560	8	75	adaptation	adaptation	NOUN
bracis-33560	8	76	 	 	SPACE
bracis-33560	9	1	[	[	X
bracis-33560	9	2	18	18	NUM
bracis-33560	9	3	]	]	PUNCT
bracis-33560	9	4	.	.	PUNCT
bracis-33560	10	1	large	large	ADJ
bracis-33560	10	2	-	-	PUNCT
bracis-33560	10	3	scale	scale	NOUN
bracis-33560	10	4	training	training	NOUN
bracis-33560	10	5	corpora	corpus	NOUN
bracis-33560	10	6	,	,	PUNCT
bracis-33560	10	7	or	or	CCONJ
bracis-33560	10	8	pre	pre	ADJ
bracis-33560	10	9	-	-	ADJ
bracis-33560	10	10	training	training	ADJ
bracis-33560	10	11	corpora	corpora	NOUN
bracis-33560	10	12	,	,	PUNCT
bracis-33560	10	13	are	be	AUX
bracis-33560	10	14	fundamental	fundamental	ADJ
bracis-33560	10	15	for	for	ADP
bracis-33560	10	16	developing	develop	VERB
bracis-33560	10	17	foundational	foundational	ADJ
bracis-33560	10	18	models	model	NOUN
bracis-33560	10	19	,	,	PUNCT
bracis-33560	10	20	which	which	PRON
bracis-33560	10	21	serve	serve	VERB
bracis-33560	10	22	as	as	ADP
bracis-33560	10	23	the	the	DET
bracis-33560	10	24	basis	basis	NOUN
bracis-33560	10	25	for	for	ADP
bracis-33560	10	26	numerous	numerous	ADJ
bracis-33560	10	27	task	task	NOUN
bracis-33560	10	28	-	-	PUNCT
bracis-33560	10	29	specific	specific	ADJ
bracis-33560	10	30	adaptations	adaptation	NOUN
bracis-33560	10	31	 	 	SPACE
bracis-33560	11	1	[	[	X
bracis-33560	11	2	7	7	NUM
bracis-33560	11	3	]	]	PUNCT
bracis-33560	11	4	.	.	PUNCT
bracis-33560	12	1	state	state	NOUN
bracis-33560	12	2	-	-	PUNCT
bracis-33560	12	3	of	of	ADP
bracis-33560	12	4	-	-	PUNCT
bracis-33560	12	5	the	the	DET
bracis-33560	12	6	-	-	PUNCT
bracis-33560	12	7	art	art	NOUN
bracis-33560	12	8	llm	llm	PROPN
bracis-33560	12	9	training	training	NOUN
bracis-33560	12	10	pipelines	pipeline	NOUN
bracis-33560	12	11	utilize	utilize	VERB
bracis-33560	12	12	various	various	ADJ
bracis-33560	12	13	types	type	NOUN
bracis-33560	12	14	of	of	ADP
bracis-33560	12	15	datasets	dataset	NOUN
bracis-33560	12	16	:	:	PUNCT
bracis-33560	12	17	(	(	PUNCT
bracis-33560	12	18	i	i	NOUN
bracis-33560	12	19	)	)	PUNCT
bracis-33560	12	20	pre	pre	ADJ
bracis-33560	12	21	-	-	NOUN
bracis-33560	12	22	training	train	VERB
bracis-33560	12	23	corpora	corpora	NOUN
bracis-33560	12	24	to	to	PART
bracis-33560	12	25	acquire	acquire	VERB
bracis-33560	12	26	language	language	NOUN
bracis-33560	12	27	structure	structure	NOUN
bracis-33560	12	28	,	,	PUNCT
bracis-33560	12	29	syntax	syntax	NOUN
bracis-33560	12	30	,	,	PUNCT
bracis-33560	12	31	and	and	CCONJ
bracis-33560	12	32	semantics	semantic	NOUN
bracis-33560	12	33	;	;	PUNCT
bracis-33560	12	34	(	(	PUNCT
bracis-33560	12	35	ii	ii	NOUN
bracis-33560	12	36	)	)	PUNCT
bracis-33560	12	37	instruction	instruction	NOUN
bracis-33560	12	38	fine	fine	ADV
bracis-33560	12	39	-	-	PUNCT
bracis-33560	12	40	tuning	tune	VERB
bracis-33560	12	41	datasets	dataset	NOUN
bracis-33560	12	42	to	to	PART
bracis-33560	12	43	enhance	enhance	VERB
bracis-33560	12	44	the	the	DET
bracis-33560	12	45	model	model	NOUN
bracis-33560	12	46	’s	’s	PART
bracis-33560	12	47	capability	capability	NOUN
bracis-33560	12	48	to	to	PART
bracis-33560	12	49	follow	follow	VERB
bracis-33560	12	50	instructions	instruction	NOUN
bracis-33560	12	51	;	;	PUNCT
bracis-33560	12	52	(	(	PUNCT
bracis-33560	12	53	iii	iii	X
bracis-33560	12	54	)	)	PUNCT
bracis-33560	12	55	preference	preference	NOUN
bracis-33560	12	56	datasets	dataset	NOUN
bracis-33560	12	57	to	to	PART
bracis-33560	12	58	rank	rank	VERB
bracis-33560	12	59	responses	response	NOUN
bracis-33560	12	60	;	;	PUNCT
bracis-33560	12	61	and	and	CCONJ
bracis-33560	12	62	(	(	PUNCT
bracis-33560	12	63	iv	iv	X
bracis-33560	12	64	)	)	PUNCT
bracis-33560	12	65	evaluation	evaluation	NOUN
bracis-33560	12	66	datasets	dataset	NOUN
bracis-33560	12	67	to	to	PART
bracis-33560	12	68	measure	measure	VERB
bracis-33560	12	69	model	model	NOUN
bracis-33560	12	70	performance	performance	NOUN
bracis-33560	12	71	 	 	SPACE
bracis-33560	13	1	[	[	X
bracis-33560	13	2	23	23	NUM
bracis-33560	13	3	]	]	PUNCT
bracis-33560	13	4	.	.	PUNCT
bracis-33560	14	1	recent	recent	ADJ
bracis-33560	14	2	research	research	NOUN
bracis-33560	14	3	shows	show	VERB
bracis-33560	14	4	that	that	SCONJ
bracis-33560	14	5	the	the	DET
bracis-33560	14	6	size	size	NOUN
bracis-33560	14	7	and	and	CCONJ
bracis-33560	14	8	diversity	diversity	NOUN
bracis-33560	14	9	of	of	ADP
bracis-33560	14	10	pre	pre	ADJ
bracis-33560	14	11	-	-	NOUN
bracis-33560	14	12	training	training	ADJ
bracis-33560	14	13	corpora	corpus	NOUN
bracis-33560	14	14	significantly	significantly	ADV
bracis-33560	14	15	impact	impact	VERB
bracis-33560	14	16	llm	llm	NOUN
bracis-33560	14	17	performance	performance	NOUN
bracis-33560	14	18	 	 	SPACE
bracis-33560	15	1	[	[	X
bracis-33560	15	2	14	14	NUM
bracis-33560	15	3	,	,	PUNCT
bracis-33560	15	4	18	18	NUM
bracis-33560	15	5	]	]	PUNCT
bracis-33560	15	6	.	.	PUNCT
bracis-33560	16	1	most	most	ADJ
bracis-33560	16	2	pre	pre	ADJ
bracis-33560	16	3	-	-	ADJ
bracis-33560	16	4	training	training	ADJ
bracis-33560	16	5	datasets	dataset	NOUN
bracis-33560	16	6	are	be	AUX
bracis-33560	16	7	available	available	ADJ
bracis-33560	16	8	in	in	ADP
bracis-33560	16	9	english	english	PROPN
bracis-33560	16	10	and	and	CCONJ
bracis-33560	16	11	chinese	chinese	PROPN
bracis-33560	16	12	,	,	PUNCT
bracis-33560	16	13	which	which	PRON
bracis-33560	16	14	are	be	AUX
bracis-33560	16	15	high	high	ADJ
bracis-33560	16	16	-	-	PUNCT
bracis-33560	16	17	resource	resource	NOUN
bracis-33560	16	18	languages	language	NOUN
bracis-33560	16	19	,	,	PUNCT
bracis-33560	16	20	while	while	SCONJ
bracis-33560	16	21	other	other	ADJ
bracis-33560	16	22	languages	language	NOUN
bracis-33560	16	23	have	have	VERB
bracis-33560	16	24	significantly	significantly	ADV
bracis-33560	16	25	fewer	few	ADJ
bracis-33560	16	26	tokens	token	NOUN
bracis-33560	16	27	 	 	SPACE
bracis-33560	17	1	[	[	X
bracis-33560	17	2	23	23	NUM
bracis-33560	17	3	]	]	PUNCT
bracis-33560	17	4	.	.	PUNCT
bracis-33560	18	1	although	although	SCONJ
bracis-33560	18	2	multilingual	multilingual	ADJ
bracis-33560	18	3	corpora	corpora	NOUN
bracis-33560	18	4	can	can	AUX
bracis-33560	18	5	help	help	VERB
bracis-33560	18	6	mitigate	mitigate	VERB
bracis-33560	18	7	data	datum	NOUN
bracis-33560	18	8	scarcity	scarcity	NOUN
bracis-33560	18	9	for	for	ADP
bracis-33560	18	10	low	low	ADJ
bracis-33560	18	11	-	-	PUNCT
bracis-33560	18	12	resource	resource	NOUN
bracis-33560	18	13	languages	language	NOUN
bracis-33560	18	14	,	,	PUNCT
bracis-33560	18	15	these	these	DET
bracis-33560	18	16	datasets	dataset	NOUN
bracis-33560	18	17	are	be	AUX
bracis-33560	18	18	often	often	ADV
bracis-33560	18	19	unbalanced	unbalanced	ADJ
bracis-33560	18	20	,	,	PUNCT
bracis-33560	18	21	favoring	favor	VERB
bracis-33560	18	22	high	high	ADJ
bracis-33560	18	23	-	-	PUNCT
bracis-33560	18	24	resource	resource	NOUN
bracis-33560	18	25	languages	language	NOUN
bracis-33560	18	26	 	 	SPACE
bracis-33560	18	27	[	[	X
bracis-33560	18	28	17	17	NUM
bracis-33560	18	29	]	]	PUNCT
bracis-33560	18	30	.	.	PUNCT
bracis-33560	19	1	this	this	DET
bracis-33560	19	2	imbalance	imbalance	NOUN
bracis-33560	19	3	affects	affect	VERB
bracis-33560	19	4	the	the	DET
bracis-33560	19	5	performance	performance	NOUN
bracis-33560	19	6	of	of	ADP
bracis-33560	19	7	multilingual	multilingual	ADJ
bracis-33560	19	8	models	model	NOUN
bracis-33560	19	9	for	for	ADP
bracis-33560	19	10	less	less	ADV
bracis-33560	19	11	-	-	PUNCT
bracis-33560	19	12	represented	represent	VERB
bracis-33560	19	13	languages	language	NOUN
bracis-33560	19	14	and	and	CCONJ
bracis-33560	19	15	models	model	NOUN
bracis-33560	19	16	trained	train	VERB
bracis-33560	19	17	on	on	ADP
bracis-33560	19	18	multilingual	multilingual	ADJ
bracis-33560	19	19	corpora	corpus	NOUN
bracis-33560	19	20	do	do	AUX
bracis-33560	19	21	not	not	PART
bracis-33560	19	22	perform	perform	VERB
bracis-33560	19	23	as	as	ADV
bracis-33560	19	24	well	well	ADV
bracis-33560	19	25	as	as	ADP
bracis-33560	19	26	those	those	PRON
bracis-33560	19	27	trained	train	VERB
bracis-33560	19	28	on	on	ADP
bracis-33560	19	29	monolingual	monolingual	ADJ
bracis-33560	19	30	corpora	corpus	NOUN
bracis-33560	19	31	 	 	SPACE
bracis-33560	20	1	[	[	X
bracis-33560	20	2	34	34	NUM
bracis-33560	20	3	]	]	PUNCT
bracis-33560	20	4	.	.	PUNCT
bracis-33560	21	1	therefore	therefore	ADV
bracis-33560	21	2	,	,	PUNCT
bracis-33560	21	3	it	it	PRON
bracis-33560	21	4	is	be	AUX
bracis-33560	21	5	essential	essential	ADJ
bracis-33560	21	6	to	to	PART
bracis-33560	21	7	train	train	VERB
bracis-33560	21	8	or	or	CCONJ
bracis-33560	21	9	fine	fine	ADJ
bracis-33560	21	10	-	-	PUNCT
bracis-33560	21	11	tune	tune	NOUN
bracis-33560	21	12	models	model	NOUN
bracis-33560	21	13	in	in	ADP
bracis-33560	21	14	the	the	DET
bracis-33560	21	15	target	target	NOUN
bracis-33560	21	16	languages	language	NOUN
bracis-33560	21	17	to	to	PART
bracis-33560	21	18	capture	capture	VERB
bracis-33560	21	19	linguistic	linguistic	ADJ
bracis-33560	21	20	nuances	nuance	NOUN
bracis-33560	21	21	,	,	PUNCT
bracis-33560	21	22	structures	structure	NOUN
bracis-33560	21	23	,	,	PUNCT
bracis-33560	21	24	and	and	CCONJ
bracis-33560	21	25	domain	domain	NOUN
bracis-33560	21	26	-	-	PUNCT
bracis-33560	21	27	specific	specific	ADJ
bracis-33560	21	28	or	or	CCONJ
bracis-33560	21	29	cultural	cultural	ADJ
bracis-33560	21	30	knowledge	knowledge	NOUN
bracis-33560	21	31	 	 	SPACE
bracis-33560	22	1	[	[	X
bracis-33560	22	2	28	28	NUM
bracis-33560	22	3	]	]	PUNCT
bracis-33560	22	4	.	.	PUNCT
bracis-33560	23	1	a	a	DET
bracis-33560	23	2	direct	direct	ADJ
bracis-33560	23	3	implication	implication	NOUN
bracis-33560	23	4	of	of	ADP
bracis-33560	23	5	this	this	DET
bracis-33560	23	6	scenario	scenario	NOUN
bracis-33560	23	7	is	be	AUX
bracis-33560	23	8	the	the	DET
bracis-33560	23	9	necessity	necessity	NOUN
bracis-33560	23	10	of	of	ADP
bracis-33560	23	11	making	make	VERB
bracis-33560	23	12	high	high	ADJ
bracis-33560	23	13	-	-	PUNCT
bracis-33560	23	14	quality	quality	NOUN
bracis-33560	23	15	plain	plain	ADJ
bracis-33560	23	16	-	-	PUNCT
bracis-33560	23	17	text	text	NOUN
bracis-33560	23	18	corpora	corpora	NOUN
bracis-33560	23	19	available	available	ADJ
bracis-33560	23	20	to	to	PART
bracis-33560	23	21	encourage	encourage	VERB
bracis-33560	23	22	research	research	NOUN
bracis-33560	23	23	on	on	ADP
bracis-33560	23	24	specific	specific	ADJ
bracis-33560	23	25	model	model	NOUN
bracis-33560	23	26	languages	language	NOUN
bracis-33560	23	27	and	and	CCONJ
bracis-33560	23	28	the	the	DET
bracis-33560	23	29	development	development	NOUN
bracis-33560	23	30	of	of	ADP
bracis-33560	23	31	better	well	ADV
bracis-33560	23	32	-	-	PUNCT
bracis-33560	23	33	performing	perform	VERB
bracis-33560	23	34	approaches	approach	NOUN
bracis-33560	23	35	.	.	PUNCT
bracis-33560	24	1	therefore	therefore	ADV
bracis-33560	24	2	,	,	PUNCT
bracis-33560	24	3	we	we	PRON
bracis-33560	24	4	introduce	introduce	VERB
bracis-33560	24	5	aroeira	aroeira	PROPN
bracis-33560	24	6	:	:	PUNCT
bracis-33560	24	7	a	a	DET
bracis-33560	24	8	curated	curate	VERB
bracis-33560	24	9	portuguese	portuguese	ADJ
bracis-33560	24	10	language	language	NOUN
bracis-33560	24	11	-	-	PUNCT
bracis-33560	24	12	specific	specific	NOUN
bracis-33560	24	13	corpus	corpus	NOUN
bracis-33560	24	14	composed	compose	VERB
bracis-33560	24	15	of	of	ADP
bracis-33560	24	16	approximately	approximately	ADV
bracis-33560	24	17	100	100	NUM
bracis-33560	24	18	gb	gb	NOUN
bracis-33560	24	19	of	of	ADP
bracis-33560	24	20	text	text	NOUN
bracis-33560	24	21	.	.	PUNCT
bracis-33560	25	1	the	the	DET
bracis-33560	25	2	content	content	NOUN
bracis-33560	25	3	was	be	AUX
bracis-33560	25	4	extracted	extract	VERB
bracis-33560	25	5	from	from	ADP
bracis-33560	25	6	recent	recent	ADJ
bracis-33560	25	7	common	common	ADJ
bracis-33560	25	8	crawlfootnote	crawlfootnote	VERB
bracis-33560	25	9	1	1	NUM
bracis-33560	25	10	(	(	PUNCT
bracis-33560	25	11	cc	cc	NOUN
bracis-33560	25	12	)	)	PUNCT
bracis-33560	25	13	web	web	NOUN
bracis-33560	25	14	pages	page	NOUN
bracis-33560	25	15	(	(	PUNCT
bracis-33560	25	16	up	up	ADP
bracis-33560	25	17	until	until	ADP
bracis-33560	25	18	2023	2023	NUM
bracis-33560	25	19	)	)	PUNCT
bracis-33560	25	20	and	and	CCONJ
bracis-33560	25	21	fully	fully	ADV
bracis-33560	25	22	curated	curate	VERB
bracis-33560	25	23	to	to	PART
bracis-33560	25	24	remove	remove	VERB
bracis-33560	25	25	web	web	NOUN
bracis-33560	25	26	tags	tag	NOUN
bracis-33560	25	27	,	,	PUNCT
bracis-33560	25	28	ensuring	ensure	VERB
bracis-33560	25	29	quality	quality	NOUN
bracis-33560	25	30	and	and	CCONJ
bracis-33560	25	31	bias	bias	NOUN
bracis-33560	25	32	filtering	filtering	NOUN
bracis-33560	25	33	.	.	PUNCT
bracis-33560	26	1	to	to	ADP
bracis-33560	26	2	the	the	DET
bracis-33560	26	3	best	good	ADJ
bracis-33560	26	4	of	of	ADP
bracis-33560	26	5	our	our	PRON
bracis-33560	26	6	knowledge	knowledge	NOUN
bracis-33560	26	7	,	,	PUNCT
bracis-33560	26	8	aroeira	aroeira	PROPN
bracis-33560	26	9	is	be	AUX
bracis-33560	26	10	the	the	DET
bracis-33560	26	11	largest	large	ADJ
bracis-33560	26	12	highly	highly	ADV
bracis-33560	26	13	-	-	PUNCT
bracis-33560	26	14	curated	curate	VERB
bracis-33560	26	15	portuguese	portuguese	ADJ
bracis-33560	26	16	corpus	corpu	NOUN
bracis-33560	26	17	available	available	ADJ
bracis-33560	26	18	to	to	ADP
bracis-33560	26	19	date	date	NOUN
bracis-33560	26	20	.	.	PUNCT
bracis-33560	27	1	it	it	PRON
bracis-33560	27	2	has	have	VERB
bracis-33560	27	3	the	the	DET
bracis-33560	27	4	potential	potential	NOUN
bracis-33560	27	5	to	to	PART
bracis-33560	27	6	influence	influence	VERB
bracis-33560	27	7	new	new	ADJ
bracis-33560	27	8	instruction	instruction	NOUN
bracis-33560	27	9	fine	fine	ADV
bracis-33560	27	10	-	-	PUNCT
bracis-33560	27	11	tuning	tuning	NOUN
bracis-33560	27	12	and	and	CCONJ
bracis-33560	27	13	evaluation	evaluation	NOUN
bracis-33560	27	14	dataset	dataset	NOUN
bracis-33560	27	15	studies	study	NOUN
bracis-33560	27	16	while	while	SCONJ
bracis-33560	27	17	guiding	guide	VERB
bracis-33560	27	18	the	the	DET
bracis-33560	27	19	development	development	NOUN
bracis-33560	27	20	of	of	ADP
bracis-33560	27	21	preference	preference	NOUN
bracis-33560	27	22	datasets	dataset	NOUN
bracis-33560	27	23	and	and	CCONJ
bracis-33560	27	24	large	large	ADJ
bracis-33560	27	25	models	model	NOUN
bracis-33560	27	26	.	.	PUNCT
bracis-33560	28	1	aroeira	aroeira	PROPN
bracis-33560	28	2	was	be	AUX
bracis-33560	28	3	created	create	VERB
bracis-33560	28	4	based	base	VERB
bracis-33560	28	5	on	on	ADP
bracis-33560	28	6	a	a	DET
bracis-33560	28	7	double	double	ADJ
bracis-33560	28	8	-	-	PUNCT
bracis-33560	28	9	pipeline	pipeline	NOUN
bracis-33560	28	10	inspired	inspire	VERB
bracis-33560	28	11	by	by	ADP
bracis-33560	28	12	 	 	SPACE
bracis-33560	28	13	[	[	X
bracis-33560	28	14	30	30	NUM
bracis-33560	28	15	]	]	PUNCT
bracis-33560	28	16	.	.	PUNCT
bracis-33560	29	1	the	the	DET
bracis-33560	29	2	pipeline	pipeline	NOUN
bracis-33560	29	3	comprises	comprise	VERB
bracis-33560	29	4	two	two	NUM
bracis-33560	29	5	key	key	ADJ
bracis-33560	29	6	steps	step	NOUN
bracis-33560	29	7	:	:	PUNCT
bracis-33560	29	8	data	datum	NOUN
bracis-33560	29	9	quality	quality	NOUN
bracis-33560	29	10	management	management	NOUN
bracis-33560	29	11	and	and	CCONJ
bracis-33560	29	12	content	content	NOUN
bracis-33560	29	13	safety	safety	NOUN
bracis-33560	29	14	assurance	assurance	NOUN
bracis-33560	29	15	.	.	PUNCT
bracis-33560	30	1	these	these	DET
bracis-33560	30	2	steps	step	NOUN
bracis-33560	30	3	ensure	ensure	VERB
bracis-33560	30	4	the	the	DET
bracis-33560	30	5	size	size	NOUN
bracis-33560	30	6	and	and	CCONJ
bracis-33560	30	7	quality	quality	NOUN
bracis-33560	30	8	necessary	necessary	ADJ
bracis-33560	30	9	for	for	SCONJ
bracis-33560	30	10	a	a	DET
bracis-33560	30	11	portuguese	portuguese	ADJ
bracis-33560	30	12	corpus	corpus	NOUN
bracis-33560	30	13	to	to	PART
bracis-33560	30	14	train	train	VERB
bracis-33560	30	15	safe	safe	ADJ
bracis-33560	30	16	llms	llm	NOUN
bracis-33560	30	17	effectively	effectively	ADV
bracis-33560	30	18	.	.	PUNCT
bracis-33560	31	1	as	as	ADP
bracis-33560	31	2	part	part	NOUN
bracis-33560	31	3	of	of	ADP
bracis-33560	31	4	the	the	DET
bracis-33560	31	5	content	content	NOUN
bracis-33560	31	6	safety	safety	NOUN
bracis-33560	31	7	step	step	NOUN
bracis-33560	31	8	,	,	PUNCT
bracis-33560	31	9	we	we	PRON
bracis-33560	31	10	investigated	investigate	VERB
bracis-33560	31	11	techniques	technique	NOUN
bracis-33560	31	12	for	for	ADP
bracis-33560	31	13	filtering	filter	VERB
bracis-33560	31	14	hazardous	hazardous	ADJ
bracis-33560	31	15	content	content	NOUN
bracis-33560	31	16	and	and	CCONJ
bracis-33560	31	17	mitigating	mitigate	VERB
bracis-33560	31	18	biases	bias	NOUN
bracis-33560	31	19	in	in	ADP
bracis-33560	31	20	our	our	PRON
bracis-33560	31	21	corpora	corpus	NOUN
bracis-33560	31	22	 	 	SPACE
bracis-33560	32	1	[	[	X
bracis-33560	32	2	16	16	NUM
bracis-33560	32	3	]	]	PUNCT
bracis-33560	32	4	.	.	PUNCT
bracis-33560	33	1	this	this	DET
bracis-33560	33	2	effort	effort	NOUN
bracis-33560	33	3	resulted	result	VERB
bracis-33560	33	4	in	in	ADP
bracis-33560	33	5	a	a	DET
bracis-33560	33	6	custom	custom	NOUN
bracis-33560	33	7	portuguese	portuguese	ADJ
bracis-33560	33	8	word	word	NOUN
bracis-33560	33	9	dictionary	dictionary	PROPN
bracis-33560	33	10	,	,	PUNCT
bracis-33560	33	11	which	which	PRON
bracis-33560	33	12	encompasses	encompass	VERB
bracis-33560	33	13	offensive	offensive	ADJ
bracis-33560	33	14	words	word	NOUN
bracis-33560	33	15	,	,	PUNCT
bracis-33560	33	16	as	as	ADV
bracis-33560	33	17	well	well	ADV
bracis-33560	33	18	as	as	ADP
bracis-33560	33	19	terms	term	NOUN
bracis-33560	33	20	,	,	PUNCT
bracis-33560	33	21	expressions	expression	NOUN
bracis-33560	33	22	,	,	PUNCT
bracis-33560	33	23	and	and	CCONJ
bracis-33560	33	24	phrases	phrase	NOUN
bracis-33560	33	25	that	that	PRON
bracis-33560	33	26	include	include	VERB
bracis-33560	33	27	sexism	sexism	NOUN
bracis-33560	33	28	,	,	PUNCT
bracis-33560	33	29	homophobia	homophobia	NOUN
bracis-33560	33	30	,	,	PUNCT
bracis-33560	33	31	ableism	ableism	NOUN
bracis-33560	33	32	,	,	PUNCT
bracis-33560	33	33	racism	racism	NOUN
bracis-33560	33	34	,	,	PUNCT
bracis-33560	33	35	hate	hate	VERB
bracis-33560	33	36	speech	speech	NOUN
bracis-33560	33	37	,	,	PUNCT
bracis-33560	33	38	and	and	CCONJ
bracis-33560	33	39	political	political	ADJ
bracis-33560	33	40	,	,	PUNCT
bracis-33560	33	41	religious	religious	ADJ
bracis-33560	33	42	,	,	PUNCT
bracis-33560	33	43	and	and	CCONJ
bracis-33560	33	44	regional	regional	ADJ
bracis-33560	33	45	prejudice	prejudice	NOUN
bracis-33560	33	46	 	 	SPACE
bracis-33560	33	47	[	[	X
bracis-33560	33	48	13	13	NUM
bracis-33560	33	49	,	,	PUNCT
bracis-33560	33	50	24	24	NUM
bracis-33560	33	51	,	,	PUNCT
bracis-33560	33	52	26	26	NUM
bracis-33560	33	53	]	]	PUNCT
bracis-33560	33	54	.	.	PUNCT
bracis-33560	34	1	we	we	PRON
bracis-33560	34	2	highlight	highlight	VERB
bracis-33560	34	3	our	our	PRON
bracis-33560	34	4	main	main	ADJ
bracis-33560	34	5	contributions	contribution	NOUN
bracis-33560	34	6	:	:	PUNCT
bracis-33560	34	7	introduction	introduction	NOUN
bracis-33560	34	8	of	of	ADP
bracis-33560	34	9	aroeira	aroeira	PROPN
bracis-33560	34	10	,	,	PUNCT
bracis-33560	34	11	a	a	DET
bracis-33560	34	12	100	100	NUM
bracis-33560	34	13	gb	gb	NOUN
bracis-33560	34	14	portuguese	portuguese	ADJ
bracis-33560	34	15	corpus	corpus	NOUN
bracis-33560	34	16	from	from	ADP
bracis-33560	34	17	diverse	diverse	ADJ
bracis-33560	34	18	internet	internet	NOUN
bracis-33560	34	19	sources	source	NOUN
bracis-33560	34	20	.	.	PUNCT
bracis-33560	35	1	our	our	PRON
bracis-33560	35	2	dataset	dataset	NOUN
bracis-33560	35	3	surpasses	surpass	VERB
bracis-33560	35	4	the	the	DET
bracis-33560	35	5	largest	large	ADJ
bracis-33560	35	6	currently	currently	ADV
bracis-33560	35	7	available	available	ADJ
bracis-33560	35	8	corpus	corpus	NOUN
bracis-33560	35	9	for	for	ADP
bracis-33560	35	10	training	training	NOUN
bracis-33560	35	11	language	language	NOUN
bracis-33560	35	12	models	model	NOUN
bracis-33560	35	13	in	in	ADP
bracis-33560	35	14	portuguese	portuguese	NOUN
bracis-33560	35	15	in	in	ADP
bracis-33560	35	16	terms	term	NOUN
bracis-33560	35	17	of	of	ADP
bracis-33560	35	18	size	size	NOUN
bracis-33560	35	19	,	,	PUNCT
bracis-33560	35	20	quality	quality	NOUN
bracis-33560	35	21	,	,	PUNCT
bracis-33560	35	22	and	and	CCONJ
bracis-33560	35	23	representativeness	representativeness	NOUN
bracis-33560	35	24	.	.	PUNCT
bracis-33560	36	1	development	development	NOUN
bracis-33560	36	2	of	of	ADP
bracis-33560	36	3	a	a	DET
bracis-33560	36	4	parameterizable	parameterizable	ADJ
bracis-33560	36	5	double	double	ADJ
bracis-33560	36	6	-	-	PUNCT
bracis-33560	36	7	pipeline	pipeline	NOUN
bracis-33560	36	8	,	,	PUNCT
bracis-33560	36	9	which	which	PRON
bracis-33560	36	10	includes	include	VERB
bracis-33560	36	11	:	:	PUNCT
bracis-33560	36	12	downloading	download	VERB
bracis-33560	36	13	,	,	PUNCT
bracis-33560	36	14	extracting	extract	VERB
bracis-33560	36	15	,	,	PUNCT
bracis-33560	36	16	language	language	NOUN
bracis-33560	36	17	identification	identification	NOUN
bracis-33560	36	18	,	,	PUNCT
bracis-33560	36	19	quality	quality	NOUN
bracis-33560	36	20	filtering	filtering	NOUN
bracis-33560	36	21	,	,	PUNCT
bracis-33560	36	22	and	and	CCONJ
bracis-33560	36	23	text	text	NOUN
bracis-33560	36	24	storage	storage	NOUN
bracis-33560	36	25	in	in	ADP
bracis-33560	36	26	the	the	DET
bracis-33560	36	27	data	data	NOUN
bracis-33560	36	28	step	step	NOUN
bracis-33560	36	29	;	;	PUNCT
bracis-33560	36	30	filtering	filter	VERB
bracis-33560	36	31	sexual	sexual	ADJ
bracis-33560	36	32	content	content	NOUN
bracis-33560	36	33	,	,	PUNCT
bracis-33560	36	34	toxic	toxic	ADJ
bracis-33560	36	35	data	datum	NOUN
bracis-33560	36	36	,	,	PUNCT
bracis-33560	36	37	and	and	CCONJ
bracis-33560	36	38	bias	bias	NOUN
bracis-33560	36	39	in	in	ADP
bracis-33560	36	40	the	the	DET
bracis-33560	36	41	content	content	NOUN
bracis-33560	36	42	safety	safety	NOUN
bracis-33560	36	43	step	step	NOUN
bracis-33560	36	44	.	.	PUNCT
bracis-33560	37	1	creation	creation	NOUN
bracis-33560	37	2	of	of	ADP
bracis-33560	37	3	a	a	DET
bracis-33560	37	4	dictionary	dictionary	NOUN
bracis-33560	37	5	to	to	PART
bracis-33560	37	6	filter	filter	VERB
bracis-33560	37	7	biased	biased	ADJ
bracis-33560	37	8	terms	term	NOUN
bracis-33560	37	9	and	and	CCONJ
bracis-33560	37	10	mitigate	mitigate	VERB
bracis-33560	37	11	social	social	ADJ
bracis-33560	37	12	bias	bias	NOUN
bracis-33560	37	13	in	in	ADP
bracis-33560	37	14	the	the	DET
bracis-33560	37	15	portuguese	portuguese	ADJ
bracis-33560	37	16	language	language	NOUN
bracis-33560	37	17	.	.	PUNCT
bracis-33560	38	1	this	this	DET
bracis-33560	38	2	paper	paper	NOUN
bracis-33560	38	3	is	be	AUX
bracis-33560	38	4	organized	organize	VERB
bracis-33560	38	5	as	as	SCONJ
bracis-33560	38	6	follows	follow	VERB
bracis-33560	38	7	.	.	PUNCT
bracis-33560	39	1	in	in	ADP
bracis-33560	39	2	sect	sect	NOUN
bracis-33560	39	3	.	.	PUNCT
bracis-33560	39	4	 	 	SPACE
bracis-33560	39	5	2	2	NUM
bracis-33560	39	6	,	,	PUNCT
bracis-33560	39	7	we	we	PRON
bracis-33560	39	8	present	present	VERB
bracis-33560	39	9	related	related	ADJ
bracis-33560	39	10	work	work	NOUN
bracis-33560	39	11	in	in	ADP
bracis-33560	39	12	corpus	corpus	NOUN
bracis-33560	39	13	extraction	extraction	NOUN
bracis-33560	39	14	.	.	PUNCT
bracis-33560	40	1	sections	section	NOUN
bracis-33560	40	2	 	 	SPACE
bracis-33560	40	3	3	3	NUM
bracis-33560	40	4	 	 	SPACE
bracis-33560	40	5	and	and	CCONJ
bracis-33560	40	6	 	 	SPACE
bracis-33560	40	7	4	4	NUM
bracis-33560	40	8	,	,	PUNCT
bracis-33560	40	9	we	we	PRON
bracis-33560	40	10	describe	describe	VERB
bracis-33560	40	11	the	the	DET
bracis-33560	40	12	methodology	methodology	NOUN
bracis-33560	40	13	for	for	ADP
bracis-33560	40	14	generating	generate	VERB
bracis-33560	40	15	the	the	DET
bracis-33560	40	16	corpus	corpus	NOUN
bracis-33560	40	17	and	and	CCONJ
bracis-33560	40	18	the	the	DET
bracis-33560	40	19	configuration	configuration	NOUN
bracis-33560	40	20	of	of	ADP
bracis-33560	40	21	hyperparameters	hyperparameter	NOUN
bracis-33560	40	22	used	use	VERB
bracis-33560	40	23	in	in	ADP
bracis-33560	40	24	the	the	DET
bracis-33560	40	25	quality	quality	NOUN
bracis-33560	40	26	filters	filter	NOUN
bracis-33560	40	27	,	,	PUNCT
bracis-33560	40	28	respectively	respectively	ADV
bracis-33560	40	29	.	.	PUNCT
bracis-33560	41	1	in	in	ADP
bracis-33560	41	2	sect	sect	NOUN
bracis-33560	41	3	.	.	PUNCT
bracis-33560	41	4	 	 	SPACE
bracis-33560	41	5	5	5	NUM
bracis-33560	41	6	,	,	PUNCT
bracis-33560	41	7	we	we	PRON
bracis-33560	41	8	analyze	analyze	VERB
bracis-33560	41	9	the	the	DET
bracis-33560	41	10	volumetry	volumetry	NOUN
bracis-33560	41	11	of	of	ADP
bracis-33560	41	12	aroeira	aroeira	PROPN
bracis-33560	41	13	in	in	ADP
bracis-33560	41	14	terms	term	NOUN
bracis-33560	41	15	of	of	ADP
bracis-33560	41	16	year	year	NOUN
bracis-33560	41	17	distribution	distribution	NOUN
bracis-33560	41	18	,	,	PUNCT
bracis-33560	41	19	knowledge	knowledge	NOUN
bracis-33560	41	20	domains	domain	NOUN
bracis-33560	41	21	,	,	PUNCT
bracis-33560	41	22	document	document	NOUN
bracis-33560	41	23	length	length	NOUN
bracis-33560	41	24	,	,	PUNCT
bracis-33560	41	25	and	and	CCONJ
bracis-33560	41	26	other	other	ADJ
bracis-33560	41	27	relevant	relevant	ADJ
bracis-33560	41	28	results	result	NOUN
bracis-33560	41	29	.	.	PUNCT
bracis-33560	42	1	finally	finally	ADV
bracis-33560	42	2	,	,	PUNCT
bracis-33560	42	3	sect	sect	NOUN
bracis-33560	42	4	.	.	PUNCT
bracis-33560	42	5	 	 	SPACE
bracis-33560	43	1	6	6	NUM
bracis-33560	43	2	presents	present	VERB
bracis-33560	43	3	conclusions	conclusion	NOUN
bracis-33560	43	4	and	and	CCONJ
bracis-33560	43	5	future	future	ADJ
bracis-33560	43	6	works	work	NOUN
bracis-33560	43	7	.	.	PUNCT
bracis-33560	44	1	2	2	NUM
bracis-33560	44	2	related	related	ADJ
bracis-33560	44	3	work	work	NOUN
bracis-33560	44	4	the	the	DET
bracis-33560	44	5	largest	large	ADJ
bracis-33560	44	6	portuguese	portuguese	ADJ
bracis-33560	44	7	language	language	NOUN
bracis-33560	44	8	corpus	corpus	NOUN
bracis-33560	44	9	is	be	AUX
bracis-33560	44	10	brwac	brwac	VERB
bracis-33560	44	11	 	 	SPACE
bracis-33560	44	12	[	[	X
bracis-33560	44	13	35	35	NUM
bracis-33560	44	14	]	]	PUNCT
bracis-33560	44	15	which	which	PRON
bracis-33560	44	16	has	have	VERB
bracis-33560	44	17	approximately	approximately	ADV
bracis-33560	44	18	25	25	NUM
bracis-33560	44	19	gb	gb	NOUN
bracis-33560	44	20	of	of	ADP
bracis-33560	44	21	textual	textual	ADJ
bracis-33560	44	22	data	datum	NOUN
bracis-33560	44	23	distributed	distribute	VERB
bracis-33560	44	24	in	in	ADP
bracis-33560	44	25	3.53	3.53	NUM
bracis-33560	44	26	mi	mi	NOUN
bracis-33560	44	27	documents	document	NOUN
bracis-33560	44	28	totaling	total	VERB
bracis-33560	44	29	2.68	2.68	NUM
bracis-33560	44	30	bi	bi	ADJ
bracis-33560	44	31	tokens	token	NOUN
bracis-33560	44	32	.	.	PUNCT
bracis-33560	45	1	another	another	DET
bracis-33560	45	2	large	large	ADJ
bracis-33560	45	3	corpus	corpus	NOUN
bracis-33560	45	4	is	be	AUX
bracis-33560	45	5	the	the	DET
bracis-33560	45	6	carolina	carolina	PROPN
bracis-33560	45	7	1.2	1.2	NUM
bracis-33560	45	8	ada	ada	PROPN
bracis-33560	46	1	[	[	X
bracis-33560	46	2	11	11	NUM
bracis-33560	46	3	]	]	PUNCT
bracis-33560	46	4	,	,	PUNCT
bracis-33560	46	5	which	which	PRON
bracis-33560	46	6	contains	contain	VERB
bracis-33560	46	7	approximately	approximately	ADV
bracis-33560	46	8	2.11	2.11	NUM
bracis-33560	46	9	mi	mi	NOUN
bracis-33560	46	10	documents	document	NOUN
bracis-33560	46	11	and	and	CCONJ
bracis-33560	46	12	a	a	DET
bracis-33560	46	13	total	total	NOUN
bracis-33560	46	14	of	of	ADP
bracis-33560	46	15	11	11	NUM
bracis-33560	46	16	gb	gb	NOUN
bracis-33560	46	17	of	of	ADP
bracis-33560	46	18	textual	textual	ADJ
bracis-33560	46	19	data	datum	NOUN
bracis-33560	46	20	.	.	PUNCT
bracis-33560	47	1	when	when	SCONJ
bracis-33560	47	2	we	we	PRON
bracis-33560	47	3	compare	compare	VERB
bracis-33560	47	4	this	this	DET
bracis-33560	47	5	corpus	corpus	NOUN
bracis-33560	47	6	with	with	ADP
bracis-33560	47	7	the	the	DET
bracis-33560	47	8	corpora	corpora	NOUN
bracis-33560	47	9	of	of	ADP
bracis-33560	47	10	other	other	ADJ
bracis-33560	47	11	languages	language	NOUN
bracis-33560	47	12	,	,	PUNCT
bracis-33560	47	13	the	the	DET
bracis-33560	47	14	gaps	gap	NOUN
bracis-33560	47	15	become	become	VERB
bracis-33560	47	16	evident	evident	ADJ
bracis-33560	47	17	.	.	PUNCT
bracis-33560	48	1	gao	gao	PROPN
bracis-33560	48	2	et	et	PROPN
bracis-33560	48	3	al	al	PROPN
bracis-33560	48	4	.	.	PUNCT
bracis-33560	48	5	 	 	SPACE
bracis-33560	49	1	[	[	X
bracis-33560	49	2	14	14	NUM
bracis-33560	49	3	]	]	PUNCT
bracis-33560	49	4	,	,	PUNCT
bracis-33560	49	5	for	for	ADP
bracis-33560	49	6	example	example	NOUN
bracis-33560	49	7	,	,	PUNCT
bracis-33560	49	8	propose	propose	VERB
bracis-33560	49	9	the	the	DET
bracis-33560	49	10	pile	pile	NOUN
bracis-33560	49	11	,	,	PUNCT
bracis-33560	49	12	a	a	DET
bracis-33560	49	13	corpus	corpus	NOUN
bracis-33560	49	14	with	with	ADP
bracis-33560	49	15	825	825	NUM
bracis-33560	49	16	gb	gb	NOUN
bracis-33560	49	17	of	of	ADP
bracis-33560	49	18	texts	text	NOUN
bracis-33560	49	19	in	in	ADP
bracis-33560	49	20	english	english	PROPN
bracis-33560	49	21	.	.	PUNCT
bracis-33560	50	1	the	the	DET
bracis-33560	50	2	corpus	corpus	NOUN
bracis-33560	50	3	is	be	AUX
bracis-33560	50	4	derived	derive	VERB
bracis-33560	50	5	from	from	ADP
bracis-33560	50	6	various	various	ADJ
bracis-33560	50	7	data	datum	NOUN
bracis-33560	50	8	sources	source	NOUN
bracis-33560	50	9	,	,	PUNCT
bracis-33560	50	10	including	include	VERB
bracis-33560	50	11	scientific	scientific	ADJ
bracis-33560	50	12	articles	article	NOUN
bracis-33560	50	13	,	,	PUNCT
bracis-33560	50	14	patent	patent	NOUN
bracis-33560	50	15	documents	document	NOUN
bracis-33560	50	16	,	,	PUNCT
bracis-33560	50	17	and	and	CCONJ
bracis-33560	50	18	forums	forum	NOUN
bracis-33560	50	19	.	.	PUNCT
bracis-33560	51	1	an	an	DET
bracis-33560	51	2	inspiring	inspiring	ADJ
bracis-33560	51	3	work	work	NOUN
bracis-33560	51	4	for	for	ADP
bracis-33560	51	5	aroeira	aroeira	PROPN
bracis-33560	51	6	is	be	AUX
bracis-33560	51	7	colossal	colossal	ADJ
bracis-33560	51	8	clean	clean	ADJ
bracis-33560	51	9	crawled	crawl	VERB
bracis-33560	51	10	corpus	corpus	NOUN
bracis-33560	51	11	(	(	PUNCT
bracis-33560	51	12	c4	c4	NOUN
bracis-33560	51	13	)	)	PUNCT
bracis-33560	51	14	 	 	SPACE
bracis-33560	52	1	[	[	X
bracis-33560	52	2	30	30	NUM
bracis-33560	52	3	]	]	PUNCT
bracis-33560	52	4	,	,	PUNCT
bracis-33560	52	5	a	a	DET
bracis-33560	52	6	curated	curate	VERB
bracis-33560	52	7	english	english	PROPN
bracis-33560	52	8	-	-	PUNCT
bracis-33560	52	9	only	only	ADV
bracis-33560	52	10	corpus	corpus	NOUN
bracis-33560	52	11	.	.	PUNCT
bracis-33560	53	1	c4	c4	NOUN
bracis-33560	53	2	was	be	AUX
bracis-33560	53	3	created	create	VERB
bracis-33560	53	4	using	use	VERB
bracis-33560	53	5	common	common	ADJ
bracis-33560	53	6	crawl	crawl	NOUN
bracis-33560	53	7	(	(	PUNCT
bracis-33560	53	8	cc	cc	NOUN
bracis-33560	53	9	)	)	PUNCT
bracis-33560	53	10	data	datum	NOUN
bracis-33560	53	11	extracted	extract	VERB
bracis-33560	53	12	in	in	ADP
bracis-33560	53	13	april	april	PROPN
bracis-33560	53	14	2019	2019	NUM
bracis-33560	53	15	and	and	CCONJ
bracis-33560	53	16	comprises	comprise	VERB
bracis-33560	53	17	approximately	approximately	ADV
bracis-33560	53	18	750	750	NUM
bracis-33560	53	19	gb	gb	NOUN
bracis-33560	53	20	of	of	ADP
bracis-33560	53	21	clean	clean	ADJ
bracis-33560	53	22	english	english	ADJ
bracis-33560	53	23	text	text	NOUN
bracis-33560	53	24	.	.	PUNCT
bracis-33560	54	1	similar	similar	ADJ
bracis-33560	54	2	to	to	ADP
bracis-33560	54	3	our	our	PRON
bracis-33560	54	4	approach	approach	NOUN
bracis-33560	54	5	,	,	PUNCT
bracis-33560	54	6	they	they	PRON
bracis-33560	54	7	apply	apply	VERB
bracis-33560	54	8	filters	filter	NOUN
bracis-33560	54	9	to	to	ADP
bracis-33560	54	10	the	the	DET
bracis-33560	54	11	raw	raw	ADJ
bracis-33560	54	12	data	datum	NOUN
bracis-33560	54	13	.	.	PUNCT
bracis-33560	55	1	cluecorpus2020	cluecorpus2020	ADP
bracis-33560	55	2	 	 	SPACE
bracis-33560	56	1	[	[	X
bracis-33560	56	2	36	36	NUM
bracis-33560	56	3	]	]	PUNCT
bracis-33560	56	4	was	be	AUX
bracis-33560	56	5	constructed	construct	VERB
bracis-33560	56	6	using	use	VERB
bracis-33560	56	7	cleaned	clean	VERB
bracis-33560	56	8	data	datum	NOUN
bracis-33560	56	9	from	from	ADP
bracis-33560	56	10	cc	cc	PROPN
bracis-33560	56	11	,	,	PUNCT
bracis-33560	56	12	resulting	result	VERB
bracis-33560	56	13	in	in	ADP
bracis-33560	56	14	a	a	DET
bracis-33560	56	15	high	high	ADJ
bracis-33560	56	16	-	-	PUNCT
bracis-33560	56	17	quality	quality	NOUN
bracis-33560	56	18	chinese	chinese	ADJ
bracis-33560	56	19	pre	pre	ADJ
bracis-33560	56	20	-	-	ADJ
bracis-33560	56	21	training	training	ADJ
bracis-33560	56	22	corpus	corpus	NOUN
bracis-33560	56	23	of	of	ADP
bracis-33560	56	24	100	100	NUM
bracis-33560	56	25	gb	gb	ADP
bracis-33560	56	26	and	and	CCONJ
bracis-33560	56	27	36	36	NUM
bracis-33560	56	28	bi	bi	ADJ
bracis-33560	56	29	tokens	token	NOUN
bracis-33560	56	30	.	.	PUNCT
bracis-33560	57	1	massivetext	massivetext	ADJ
bracis-33560	57	2	 	 	SPACE
bracis-33560	57	3	[	[	X
bracis-33560	57	4	29	29	NUM
bracis-33560	57	5	]	]	PUNCT
bracis-33560	57	6	is	be	AUX
bracis-33560	57	7	a	a	DET
bracis-33560	57	8	collection	collection	NOUN
bracis-33560	57	9	of	of	ADP
bracis-33560	57	10	large	large	ADJ
bracis-33560	57	11	english	english	ADJ
bracis-33560	57	12	datasets	dataset	NOUN
bracis-33560	57	13	created	create	VERB
bracis-33560	57	14	with	with	ADP
bracis-33560	57	15	data	datum	NOUN
bracis-33560	57	16	from	from	ADP
bracis-33560	57	17	different	different	ADJ
bracis-33560	57	18	sources	source	NOUN
bracis-33560	57	19	.	.	PUNCT
bracis-33560	58	1	massivetext	massivetext	NOUN
bracis-33560	58	2	contains	contain	VERB
bracis-33560	58	3	2.35	2.35	NUM
bracis-33560	58	4	bi	bi	ADJ
bracis-33560	58	5	documents	document	NOUN
bracis-33560	58	6	,	,	PUNCT
bracis-33560	58	7	equivalent	equivalent	ADJ
bracis-33560	58	8	to	to	ADP
bracis-33560	58	9	10.5	10.5	NUM
bracis-33560	58	10	tb	tb	NOUN
bracis-33560	58	11	of	of	ADP
bracis-33560	58	12	text	text	NOUN
bracis-33560	58	13	.	.	PUNCT
bracis-33560	59	1	more	more	ADV
bracis-33560	59	2	recently	recently	ADV
bracis-33560	59	3	,	,	PUNCT
bracis-33560	59	4	sabiá	sabiá	NOUN
bracis-33560	59	5	 	 	SPACE
bracis-33560	60	1	[	[	X
bracis-33560	60	2	28	28	NUM
bracis-33560	60	3	]	]	PUNCT
bracis-33560	60	4	applied	apply	VERB
bracis-33560	60	5	a	a	DET
bracis-33560	60	6	similar	similar	ADJ
bracis-33560	60	7	filtering	filtering	NOUN
bracis-33560	60	8	methodology	methodology	NOUN
bracis-33560	60	9	of	of	ADP
bracis-33560	60	10	massivetext	massivetext	NOUN
bracis-33560	60	11	to	to	ADP
bracis-33560	60	12	the	the	DET
bracis-33560	60	13	portuguese	portuguese	ADJ
bracis-33560	60	14	section	section	NOUN
bracis-33560	60	15	of	of	ADP
bracis-33560	60	16	clueweb	clueweb	PROPN
bracis-33560	60	17	dataset	dataset	PROPN
bracis-33560	60	18	 	 	SPACE
bracis-33560	61	1	[	[	X
bracis-33560	61	2	27	27	NUM
bracis-33560	61	3	]	]	PUNCT
bracis-33560	61	4	and	and	CCONJ
bracis-33560	61	5	managed	manage	VERB
bracis-33560	61	6	to	to	PART
bracis-33560	61	7	retrieve	retrieve	VERB
bracis-33560	61	8	a	a	DET
bracis-33560	61	9	curated	curate	VERB
bracis-33560	61	10	dataset	dataset	NOUN
bracis-33560	61	11	.	.	PUNCT
bracis-33560	62	1	wudaocorpora	wudaocorpora	PROPN
bracis-33560	62	2	 	 	SPACE
bracis-33560	63	1	[	[	X
bracis-33560	63	2	38	38	NUM
bracis-33560	63	3	]	]	PUNCT
bracis-33560	63	4	is	be	AUX
bracis-33560	63	5	a	a	DET
bracis-33560	63	6	3	3	NUM
bracis-33560	63	7	tb	tb	ADP
bracis-33560	63	8	chinese	chinese	ADJ
bracis-33560	63	9	corpus	corpus	NOUN
bracis-33560	63	10	with	with	ADP
bracis-33560	63	11	1.08	1.08	NUM
bracis-33560	63	12	tri	tri	NOUN
bracis-33560	63	13	of	of	ADP
bracis-33560	63	14	hanzi	hanzi	NOUN
bracis-33560	63	15	characters	character	NOUN
bracis-33560	63	16	collected	collect	VERB
bracis-33560	63	17	from	from	ADP
bracis-33560	63	18	822	822	NUM
bracis-33560	63	19	mi	mi	PROPN
bracis-33560	63	20	web	web	NOUN
bracis-33560	63	21	pages	page	NOUN
bracis-33560	63	22	.	.	PUNCT
bracis-33560	64	1	it	it	PRON
bracis-33560	64	2	is	be	AUX
bracis-33560	64	3	also	also	ADV
bracis-33560	64	4	worth	worth	ADJ
bracis-33560	64	5	mentioning	mention	VERB
bracis-33560	64	6	that	that	SCONJ
bracis-33560	64	7	a	a	DET
bracis-33560	64	8	current	current	ADJ
bracis-33560	64	9	trend	trend	NOUN
bracis-33560	64	10	is	be	AUX
bracis-33560	64	11	the	the	DET
bracis-33560	64	12	proposal	proposal	NOUN
bracis-33560	64	13	of	of	ADP
bracis-33560	64	14	multilingual	multilingual	ADJ
bracis-33560	64	15	corpora	corpus	NOUN
bracis-33560	64	16	.	.	PUNCT
bracis-33560	65	1	the	the	DET
bracis-33560	65	2	c4corpus	c4corpus	PROPN
bracis-33560	65	3	authors	author	NOUN
bracis-33560	65	4	 	 	SPACE
bracis-33560	65	5	[	[	X
bracis-33560	65	6	17	17	NUM
bracis-33560	65	7	]	]	PUNCT
bracis-33560	65	8	present	present	VERB
bracis-33560	65	9	the	the	DET
bracis-33560	65	10	construction	construction	NOUN
bracis-33560	65	11	of	of	ADP
bracis-33560	65	12	a	a	DET
bracis-33560	65	13	12	12	NUM
bracis-33560	65	14	mi	mi	NOUN
bracis-33560	65	15	web	web	NOUN
bracis-33560	65	16	page	page	NOUN
bracis-33560	65	17	corpus	corpus	NOUN
bracis-33560	65	18	containing	contain	VERB
bracis-33560	65	19	more	more	ADJ
bracis-33560	65	20	than	than	ADP
bracis-33560	65	21	50	50	NUM
bracis-33560	65	22	languages	language	NOUN
bracis-33560	65	23	,	,	PUNCT
bracis-33560	65	24	including	include	VERB
bracis-33560	65	25	portuguese	portuguese	PROPN
bracis-33560	65	26	.	.	PUNCT
bracis-33560	66	1	english	english	PROPN
bracis-33560	66	2	has	have	VERB
bracis-33560	66	3	a	a	DET
bracis-33560	66	4	volume	volume	NOUN
bracis-33560	66	5	of	of	ADP
bracis-33560	66	6	7.7	7.7	NUM
bracis-33560	66	7	mi	mi	NOUN
bracis-33560	66	8	(	(	PUNCT
bracis-33560	66	9	64.2	64.2	NUM
bracis-33560	66	10	%	%	NOUN
bracis-33560	66	11	)	)	PUNCT
bracis-33560	66	12	documents	document	NOUN
bracis-33560	66	13	while	while	SCONJ
bracis-33560	66	14	portuguese	portuguese	PROPN
bracis-33560	66	15	has	have	VERB
bracis-33560	66	16	only	only	ADV
bracis-33560	66	17	0.3	0.3	NUM
bracis-33560	66	18	mi	mi	X
bracis-33560	66	19	(	(	PUNCT
bracis-33560	66	20	2.5	2.5	NUM
bracis-33560	66	21	%	%	NOUN
bracis-33560	66	22	)	)	PUNCT
bracis-33560	66	23	.	.	PUNCT
bracis-33560	67	1	redpajama	redpajama	NOUN
bracis-33560	67	2	 	 	SPACE
bracis-33560	68	1	[	[	X
bracis-33560	68	2	10	10	NUM
bracis-33560	68	3	]	]	PUNCT
bracis-33560	68	4	is	be	AUX
bracis-33560	68	5	a	a	DET
bracis-33560	68	6	large	large	ADJ
bracis-33560	68	7	multilingual	multilingual	ADJ
bracis-33560	68	8	corpus	corpus	NOUN
bracis-33560	68	9	,	,	PUNCT
bracis-33560	68	10	containing	contain	VERB
bracis-33560	68	11	100	100	NUM
bracis-33560	68	12	bi	bi	ADJ
bracis-33560	68	13	text	text	NOUN
bracis-33560	68	14	documents	document	NOUN
bracis-33560	68	15	extracted	extract	VERB
bracis-33560	68	16	from	from	ADP
bracis-33560	68	17	84	84	NUM
bracis-33560	68	18	cc	cc	NOUN
bracis-33560	68	19	snapshots	snapshot	NOUN
bracis-33560	68	20	.	.	PUNCT
bracis-33560	69	1	quality	quality	NOUN
bracis-33560	69	2	signals	signal	NOUN
bracis-33560	69	3	were	be	AUX
bracis-33560	69	4	applied	apply	VERB
bracis-33560	69	5	to	to	ADP
bracis-33560	69	6	30	30	NUM
bracis-33560	69	7	billion	billion	NUM
bracis-33560	69	8	documents	document	NOUN
bracis-33560	69	9	,	,	PUNCT
bracis-33560	69	10	and	and	CCONJ
bracis-33560	69	11	deduplication	deduplication	NOUN
bracis-33560	69	12	was	be	AUX
bracis-33560	69	13	performed	perform	VERB
bracis-33560	69	14	on	on	ADP
bracis-33560	69	15	20	20	NUM
bracis-33560	69	16	billion	billion	NUM
bracis-33560	69	17	documents	document	NOUN
bracis-33560	69	18	.	.	PUNCT
bracis-33560	70	1	it	it	PRON
bracis-33560	70	2	claims	claim	VERB
bracis-33560	70	3	to	to	PART
bracis-33560	70	4	have	have	VERB
bracis-33560	70	5	english	english	PROPN
bracis-33560	70	6	(	(	PUNCT
bracis-33560	70	7	69.8	69.8	NUM
bracis-33560	70	8	%	%	NOUN
bracis-33560	70	9	)	)	PUNCT
bracis-33560	70	10	,	,	PUNCT
bracis-33560	70	11	deutch	deutch	VERB
bracis-33560	70	12	(	(	PUNCT
bracis-33560	70	13	9.2	9.2	NUM
bracis-33560	70	14	%	%	NOUN
bracis-33560	70	15	)	)	PUNCT
bracis-33560	70	16	,	,	PUNCT
bracis-33560	70	17	spanish	spanish	NOUN
bracis-33560	70	18	(	(	PUNCT
bracis-33560	70	19	8.8	8.8	NUM
bracis-33560	70	20	%	%	NOUN
bracis-33560	70	21	)	)	PUNCT
bracis-33560	70	22	,	,	PUNCT
bracis-33560	70	23	french	french	NOUN
bracis-33560	70	24	(	(	PUNCT
bracis-33560	70	25	7.8	7.8	NUM
bracis-33560	70	26	%	%	NOUN
bracis-33560	70	27	)	)	PUNCT
bracis-33560	70	28	,	,	PUNCT
bracis-33560	70	29	and	and	CCONJ
bracis-33560	70	30	italian	italian	ADJ
bracis-33560	70	31	(	(	PUNCT
bracis-33560	70	32	4.4	4.4	NUM
bracis-33560	70	33	%	%	NOUN
bracis-33560	70	34	)	)	PUNCT
bracis-33560	70	35	.	.	PUNCT
bracis-33560	71	1	as	as	SCONJ
bracis-33560	71	2	we	we	PRON
bracis-33560	71	3	can	can	AUX
bracis-33560	71	4	see	see	VERB
bracis-33560	71	5	,	,	PUNCT
bracis-33560	71	6	the	the	DET
bracis-33560	71	7	corpora	corpora	NOUN
bracis-33560	71	8	available	available	ADJ
bracis-33560	71	9	in	in	ADP
bracis-33560	71	10	english	english	PROPN
bracis-33560	71	11	and	and	CCONJ
bracis-33560	71	12	chinese	chinese	PROPN
bracis-33560	71	13	have	have	VERB
bracis-33560	71	14	a	a	DET
bracis-33560	71	15	massive	massive	ADJ
bracis-33560	71	16	data	data	NOUN
bracis-33560	71	17	amount	amount	NOUN
bracis-33560	71	18	,	,	PUNCT
bracis-33560	71	19	easily	easily	ADV
bracis-33560	71	20	surpassing	surpass	VERB
bracis-33560	71	21	corpora	corpora	NOUN
bracis-33560	71	22	in	in	ADP
bracis-33560	71	23	portuguese	portuguese	ADJ
bracis-33560	71	24	and	and	CCONJ
bracis-33560	71	25	other	other	ADJ
bracis-33560	71	26	languages	language	NOUN
bracis-33560	71	27	.	.	PUNCT
bracis-33560	72	1	however	however	ADV
bracis-33560	72	2	,	,	PUNCT
bracis-33560	72	3	we	we	PRON
bracis-33560	72	4	know	know	VERB
bracis-33560	72	5	that	that	SCONJ
bracis-33560	72	6	the	the	DET
bracis-33560	72	7	amount	amount	NOUN
bracis-33560	72	8	of	of	ADP
bracis-33560	72	9	internet	internet	NOUN
bracis-33560	72	10	information	information	NOUN
bracis-33560	72	11	available	available	ADJ
bracis-33560	72	12	in	in	ADP
bracis-33560	72	13	english	english	PROPN
bracis-33560	72	14	and	and	CCONJ
bracis-33560	72	15	chinese	chinese	PROPN
bracis-33560	72	16	is	be	AUX
bracis-33560	72	17	greater	great	ADJ
bracis-33560	72	18	than	than	ADP
bracis-33560	72	19	in	in	ADP
bracis-33560	72	20	portuguese	portuguese	PROPN
bracis-33560	72	21	.	.	PUNCT
bracis-33560	73	1	another	another	DET
bracis-33560	73	2	important	important	ADJ
bracis-33560	73	3	aspect	aspect	NOUN
bracis-33560	73	4	is	be	AUX
bracis-33560	73	5	the	the	DET
bracis-33560	73	6	biases	bias	NOUN
bracis-33560	73	7	present	present	ADJ
bracis-33560	73	8	in	in	ADP
bracis-33560	73	9	corpora	corpora	NOUN
bracis-33560	73	10	and	and	CCONJ
bracis-33560	73	11	texts	text	NOUN
bracis-33560	73	12	.	.	PUNCT
bracis-33560	74	1	language	language	NOUN
bracis-33560	74	2	is	be	AUX
bracis-33560	74	3	a	a	DET
bracis-33560	74	4	highly	highly	ADV
bracis-33560	74	5	relevant	relevant	ADJ
bracis-33560	74	6	avenue	avenue	NOUN
bracis-33560	74	7	for	for	ADP
bracis-33560	74	8	manifesting	manifest	VERB
bracis-33560	74	9	social	social	ADJ
bracis-33560	74	10	hierarchies	hierarchy	NOUN
bracis-33560	74	11	,	,	PUNCT
bracis-33560	74	12	pre	pre	ADJ
bracis-33560	74	13	-	-	ADJ
bracis-33560	74	14	established	establish	VERB
bracis-33560	74	15	concepts	concept	NOUN
bracis-33560	74	16	,	,	PUNCT
bracis-33560	74	17	and	and	CCONJ
bracis-33560	74	18	standard	standard	ADJ
bracis-33560	74	19	forms	form	NOUN
bracis-33560	74	20	of	of	ADP
bracis-33560	74	21	treatment	treatment	NOUN
bracis-33560	74	22	 	 	SPACE
bracis-33560	75	1	[	[	X
bracis-33560	75	2	6	6	NUM
bracis-33560	75	3	]	]	PUNCT
bracis-33560	75	4	.	.	PUNCT
bracis-33560	76	1	various	various	ADJ
bracis-33560	76	2	efforts	effort	NOUN
bracis-33560	76	3	are	be	AUX
bracis-33560	76	4	being	be	AUX
bracis-33560	76	5	made	make	VERB
bracis-33560	76	6	to	to	PART
bracis-33560	76	7	evaluate	evaluate	VERB
bracis-33560	76	8	data	datum	NOUN
bracis-33560	76	9	biases	bias	NOUN
bracis-33560	76	10	and	and	CCONJ
bracis-33560	76	11	how	how	SCONJ
bracis-33560	76	12	they	they	PRON
bracis-33560	76	13	impact	impact	VERB
bracis-33560	76	14	the	the	DET
bracis-33560	76	15	behavior	behavior	NOUN
bracis-33560	76	16	of	of	ADP
bracis-33560	76	17	language	language	NOUN
bracis-33560	76	18	models	model	NOUN
bracis-33560	76	19	.	.	PUNCT
bracis-33560	77	1	the	the	DET
bracis-33560	77	2	paper	paper	NOUN
bracis-33560	77	3	 	 	SPACE
bracis-33560	78	1	[	[	X
bracis-33560	78	2	24	24	NUM
bracis-33560	78	3	]	]	PUNCT
bracis-33560	78	4	analyzed	analyze	VERB
bracis-33560	78	5	93	93	NUM
bracis-33560	78	6	social	social	ADJ
bracis-33560	78	7	groups	group	NOUN
bracis-33560	78	8	that	that	PRON
bracis-33560	78	9	receive	receive	VERB
bracis-33560	78	10	stigmatized	stigmatize	VERB
bracis-33560	78	11	treatment	treatment	NOUN
bracis-33560	78	12	by	by	ADP
bracis-33560	78	13	nlp	nlp	NOUN
bracis-33560	78	14	models	model	NOUN
bracis-33560	78	15	.	.	PUNCT
bracis-33560	79	1	work	work	NOUN
bracis-33560	79	2	 	 	SPACE
bracis-33560	80	1	[	[	X
bracis-33560	80	2	25	25	NUM
bracis-33560	80	3	]	]	PUNCT
bracis-33560	80	4	created	create	VERB
bracis-33560	80	5	stereoset	stereoset	NOUN
bracis-33560	80	6	to	to	PART
bracis-33560	80	7	measure	measure	VERB
bracis-33560	80	8	stereotypical	stereotypical	ADJ
bracis-33560	80	9	treatment	treatment	NOUN
bracis-33560	80	10	in	in	ADP
bracis-33560	80	11	certain	certain	ADJ
bracis-33560	80	12	ethnic	ethnic	ADJ
bracis-33560	80	13	groups	group	NOUN
bracis-33560	80	14	,	,	PUNCT
bracis-33560	80	15	and	and	CCONJ
bracis-33560	80	16	paper	paper	NOUN
bracis-33560	80	17	 	 	SPACE
bracis-33560	81	1	[	[	X
bracis-33560	81	2	26	26	NUM
bracis-33560	81	3	]	]	PUNCT
bracis-33560	81	4	developed	develop	VERB
bracis-33560	81	5	a	a	DET
bracis-33560	81	6	benchmark	benchmark	NOUN
bracis-33560	81	7	dataset	dataset	NOUN
bracis-33560	81	8	for	for	ADP
bracis-33560	81	9	measuring	measure	VERB
bracis-33560	81	10	biases	bias	NOUN
bracis-33560	81	11	related	relate	VERB
bracis-33560	81	12	to	to	ADP
bracis-33560	81	13	gender	gender	NOUN
bracis-33560	81	14	,	,	PUNCT
bracis-33560	81	15	race	race	NOUN
bracis-33560	81	16	,	,	PUNCT
bracis-33560	81	17	age	age	NOUN
bracis-33560	81	18	,	,	PUNCT
bracis-33560	81	19	sexual	sexual	ADJ
bracis-33560	81	20	orientation	orientation	NOUN
bracis-33560	81	21	,	,	PUNCT
bracis-33560	81	22	and	and	CCONJ
bracis-33560	81	23	others	other	NOUN
bracis-33560	81	24	.	.	PUNCT
bracis-33560	82	1	these	these	DET
bracis-33560	82	2	aspects	aspect	NOUN
bracis-33560	82	3	are	be	AUX
bracis-33560	82	4	relevant	relevant	ADJ
bracis-33560	82	5	in	in	ADP
bracis-33560	82	6	a	a	DET
bracis-33560	82	7	context	context	NOUN
bracis-33560	82	8	with	with	ADP
bracis-33560	82	9	strong	strong	ADJ
bracis-33560	82	10	normative	normative	ADJ
bracis-33560	82	11	motivations	motivation	NOUN
bracis-33560	82	12	and	and	CCONJ
bracis-33560	82	13	the	the	DET
bracis-33560	82	14	need	need	NOUN
bracis-33560	82	15	to	to	PART
bracis-33560	82	16	create	create	VERB
bracis-33560	82	17	responsible	responsible	ADJ
bracis-33560	82	18	ai	ai	NOUN
bracis-33560	82	19	.	.	PUNCT
bracis-33560	83	1	many	many	ADJ
bracis-33560	83	2	ways	way	NOUN
bracis-33560	83	3	exist	exist	VERB
bracis-33560	83	4	to	to	PART
bracis-33560	83	5	mitigate	mitigate	VERB
bracis-33560	83	6	text	text	NOUN
bracis-33560	83	7	biases	bias	NOUN
bracis-33560	83	8	,	,	PUNCT
bracis-33560	83	9	such	such	ADJ
bracis-33560	83	10	as	as	ADP
bracis-33560	83	11	data	data	NOUN
bracis-33560	83	12	augmentation	augmentation	NOUN
bracis-33560	83	13	,	,	PUNCT
bracis-33560	83	14	content	content	NOUN
bracis-33560	83	15	filtering	filtering	NOUN
bracis-33560	83	16	,	,	PUNCT
bracis-33560	83	17	rebalancing	rebalancing	NOUN
bracis-33560	83	18	,	,	PUNCT
bracis-33560	83	19	masking	masking	NOUN
bracis-33560	83	20	,	,	PUNCT
bracis-33560	83	21	and	and	CCONJ
bracis-33560	83	22	many	many	ADJ
bracis-33560	83	23	others	other	NOUN
bracis-33560	83	24	 	 	SPACE
bracis-33560	84	1	[	[	X
bracis-33560	84	2	13	13	NUM
bracis-33560	84	3	]	]	PUNCT
bracis-33560	84	4	.	.	PUNCT
bracis-33560	85	1	our	our	PRON
bracis-33560	85	2	work	work	NOUN
bracis-33560	85	3	uses	use	VERB
bracis-33560	85	4	the	the	DET
bracis-33560	85	5	concept	concept	NOUN
bracis-33560	85	6	studied	study	VERB
bracis-33560	85	7	by	by	ADP
bracis-33560	85	8	 	 	SPACE
bracis-33560	85	9	[	[	X
bracis-33560	85	10	16	16	NUM
bracis-33560	85	11	]	]	PUNCT
bracis-33560	85	12	,	,	PUNCT
bracis-33560	85	13	where	where	SCONJ
bracis-33560	85	14	filtering	filter	VERB
bracis-33560	85	15	sensitive	sensitive	ADJ
bracis-33560	85	16	content	content	NOUN
bracis-33560	85	17	can	can	AUX
bracis-33560	85	18	result	result	VERB
bracis-33560	85	19	in	in	ADP
bracis-33560	85	20	models	model	NOUN
bracis-33560	85	21	with	with	ADP
bracis-33560	85	22	more	more	ADV
bracis-33560	85	23	equitable	equitable	ADJ
bracis-33560	85	24	treatment	treatment	NOUN
bracis-33560	85	25	of	of	ADP
bracis-33560	85	26	different	different	ADJ
bracis-33560	85	27	ethnic	ethnic	ADJ
bracis-33560	85	28	groups	group	NOUN
bracis-33560	85	29	.	.	PUNCT
bracis-33560	86	1	the	the	DET
bracis-33560	86	2	work	work	NOUN
bracis-33560	86	3	specifically	specifically	ADV
bracis-33560	86	4	uses	use	VERB
bracis-33560	86	5	word	word	NOUN
bracis-33560	86	6	co	co	NOUN
bracis-33560	86	7	-	-	NOUN
bracis-33560	86	8	occurrence	occurrence	NOUN
bracis-33560	86	9	in	in	ADP
bracis-33560	86	10	the	the	DET
bracis-33560	86	11	filtering	filtering	NOUN
bracis-33560	86	12	process	process	NOUN
bracis-33560	86	13	.	.	PUNCT
bracis-33560	87	1	based	base	VERB
bracis-33560	87	2	on	on	ADP
bracis-33560	87	3	these	these	DET
bracis-33560	87	4	past	past	ADJ
bracis-33560	87	5	works	work	NOUN
bracis-33560	87	6	,	,	PUNCT
bracis-33560	87	7	we	we	PRON
bracis-33560	87	8	can	can	AUX
bracis-33560	87	9	see	see	VERB
bracis-33560	87	10	that	that	PRON
bracis-33560	87	11	,	,	PUNCT
bracis-33560	87	12	in	in	ADP
bracis-33560	87	13	general	general	ADJ
bracis-33560	87	14	,	,	PUNCT
bracis-33560	87	15	they	they	PRON
bracis-33560	87	16	focus	focus	VERB
bracis-33560	87	17	on	on	ADP
bracis-33560	87	18	creating	create	VERB
bracis-33560	87	19	corpora	corpora	NOUN
bracis-33560	87	20	for	for	ADP
bracis-33560	87	21	training	training	NOUN
bracis-33560	87	22	language	language	NOUN
bracis-33560	87	23	models	model	NOUN
bracis-33560	87	24	for	for	ADP
bracis-33560	87	25	high	high	ADJ
bracis-33560	87	26	-	-	PUNCT
bracis-33560	87	27	resource	resource	NOUN
bracis-33560	87	28	language	language	NOUN
bracis-33560	87	29	tasks	task	NOUN
bracis-33560	87	30	.	.	PUNCT
bracis-33560	88	1	thus	thus	ADV
bracis-33560	88	2	,	,	PUNCT
bracis-33560	88	3	there	there	PRON
bracis-33560	88	4	is	be	VERB
bracis-33560	88	5	a	a	DET
bracis-33560	88	6	necessity	necessity	NOUN
bracis-33560	88	7	for	for	ADP
bracis-33560	88	8	creating	create	VERB
bracis-33560	88	9	a	a	DET
bracis-33560	88	10	portuguese	portuguese	ADJ
bracis-33560	88	11	corpus	corpus	NOUN
bracis-33560	88	12	since	since	SCONJ
bracis-33560	88	13	the	the	DET
bracis-33560	88	14	amount	amount	NOUN
bracis-33560	88	15	of	of	ADP
bracis-33560	88	16	large	large	ADJ
bracis-33560	88	17	brazilian	brazilian	ADJ
bracis-33560	88	18	portuguese	portuguese	ADJ
bracis-33560	88	19	models	model	NOUN
bracis-33560	88	20	has	have	AUX
bracis-33560	88	21	drastically	drastically	ADV
bracis-33560	88	22	increased	increase	VERB
bracis-33560	88	23	recently	recently	ADV
bracis-33560	88	24	.	.	PUNCT
bracis-33560	89	1	we	we	PRON
bracis-33560	89	2	can	can	AUX
bracis-33560	89	3	cite	cite	VERB
bracis-33560	89	4	bertimbau	bertimbau	VERB
bracis-33560	89	5	 	 	SPACE
bracis-33560	90	1	[	[	X
bracis-33560	90	2	33	33	NUM
bracis-33560	90	3	]	]	PUNCT
bracis-33560	90	4	,	,	PUNCT
bracis-33560	90	5	ptt5	ptt5	PROPN
bracis-33560	90	6	 	 	SPACE
bracis-33560	91	1	[	[	X
bracis-33560	91	2	9	9	NUM
bracis-33560	91	3	]	]	PUNCT
bracis-33560	91	4	,	,	PUNCT
bracis-33560	91	5	bertaú	bertaú	PROPN
bracis-33560	91	6	 	 	SPACE
bracis-33560	92	1	[	[	X
bracis-33560	92	2	12	12	NUM
bracis-33560	92	3	]	]	PUNCT
bracis-33560	92	4	,	,	PUNCT
bracis-33560	92	5	sabiá	sabiá	PROPN
bracis-33560	92	6	 	 	SPACE
bracis-33560	92	7	[	[	X
bracis-33560	92	8	4	4	NUM
bracis-33560	92	9	,	,	PUNCT
bracis-33560	92	10	28	28	NUM
bracis-33560	92	11	]	]	PUNCT
bracis-33560	92	12	,	,	PUNCT
bracis-33560	92	13	cabrita	cabrita	PROPN
bracis-33560	92	14	 	 	SPACE
bracis-33560	93	1	[	[	X
bracis-33560	93	2	22	22	NUM
bracis-33560	93	3	]	]	PUNCT
bracis-33560	93	4	,	,	PUNCT
bracis-33560	94	1	and	and	CCONJ
bracis-33560	94	2	bode	bode	VERB
bracis-33560	94	3	 	 	SPACE
bracis-33560	94	4	[	[	X
bracis-33560	94	5	15	15	NUM
bracis-33560	94	6	]	]	PUNCT
bracis-33560	94	7	.	.	PUNCT
bracis-33560	95	1	3	3	NUM
bracis-33560	95	2	aroeira	aroeira	NOUN
bracis-33560	95	3	in	in	ADP
bracis-33560	95	4	this	this	DET
bracis-33560	95	5	section	section	NOUN
bracis-33560	95	6	,	,	PUNCT
bracis-33560	95	7	we	we	PRON
bracis-33560	95	8	detail	detail	VERB
bracis-33560	95	9	the	the	DET
bracis-33560	95	10	steps	step	NOUN
bracis-33560	95	11	of	of	ADP
bracis-33560	95	12	the	the	DET
bracis-33560	95	13	corpus	corpus	NOUN
bracis-33560	95	14	creation	creation	NOUN
bracis-33560	95	15	(	(	PUNCT
bracis-33560	95	16	double	double	ADJ
bracis-33560	95	17	-	-	PUNCT
bracis-33560	95	18	pipeline	pipeline	NOUN
bracis-33560	95	19	)	)	PUNCT
bracis-33560	95	20	which	which	PRON
bracis-33560	95	21	is	be	AUX
bracis-33560	95	22	divided	divide	VERB
bracis-33560	95	23	into	into	ADP
bracis-33560	95	24	two	two	NUM
bracis-33560	95	25	objectives	objective	NOUN
bracis-33560	95	26	,	,	PUNCT
bracis-33560	95	27	(	(	PUNCT
bracis-33560	95	28	i	i	NOUN
bracis-33560	95	29	)	)	PUNCT
bracis-33560	95	30	collect	collect	NOUN
bracis-33560	95	31	(	(	PUNCT
bracis-33560	95	32	data	data	NOUN
bracis-33560	95	33	pipeline	pipeline	NOUN
bracis-33560	95	34	)	)	PUNCT
bracis-33560	95	35	and	and	CCONJ
bracis-33560	95	36	(	(	PUNCT
bracis-33560	95	37	ii	ii	NOUN
bracis-33560	95	38	)	)	PUNCT
bracis-33560	95	39	ensure	ensure	VERB
bracis-33560	95	40	content	content	NOUN
bracis-33560	95	41	safety	safety	NOUN
bracis-33560	95	42	(	(	PUNCT
bracis-33560	95	43	content	content	NOUN
bracis-33560	95	44	safety	safety	NOUN
bracis-33560	95	45	pipeline	pipeline	PROPN
bracis-33560	95	46	)	)	PUNCT
bracis-33560	95	47	.	.	PUNCT
bracis-33560	96	1	our	our	PRON
bracis-33560	96	2	whole	whole	ADJ
bracis-33560	96	3	pipeline	pipeline	NOUN
bracis-33560	96	4	contains	contain	VERB
bracis-33560	96	5	nine	nine	NUM
bracis-33560	96	6	steps	step	NOUN
bracis-33560	96	7	:	:	PUNCT
bracis-33560	96	8	data	datum	NOUN
bracis-33560	96	9	collection	collection	NOUN
bracis-33560	96	10	and	and	CCONJ
bracis-33560	96	11	sampling	sampling	NOUN
bracis-33560	96	12	,	,	PUNCT
bracis-33560	96	13	text	text	NOUN
bracis-33560	96	14	extraction	extraction	NOUN
bracis-33560	96	15	,	,	PUNCT
bracis-33560	96	16	language	language	NOUN
bracis-33560	96	17	identification	identification	NOUN
bracis-33560	96	18	,	,	PUNCT
bracis-33560	96	19	deduplication	deduplication	NOUN
bracis-33560	96	20	,	,	PUNCT
bracis-33560	96	21	and	and	CCONJ
bracis-33560	96	22	quality	quality	NOUN
bracis-33560	96	23	filters	filter	NOUN
bracis-33560	96	24	in	in	ADP
bracis-33560	96	25	data	datum	NOUN
bracis-33560	96	26	pipeline	pipeline	NOUN
bracis-33560	96	27	,	,	PUNCT
bracis-33560	96	28	and	and	CCONJ
bracis-33560	96	29	sexual	sexual	ADJ
bracis-33560	96	30	content	content	NOUN
bracis-33560	96	31	filter	filter	NOUN
bracis-33560	96	32	,	,	PUNCT
bracis-33560	96	33	toxic	toxic	ADJ
bracis-33560	96	34	data	datum	NOUN
bracis-33560	96	35	filter	filter	NOUN
bracis-33560	96	36	,	,	PUNCT
bracis-33560	96	37	bias	bias	NOUN
bracis-33560	96	38	filter	filter	NOUN
bracis-33560	96	39	,	,	PUNCT
bracis-33560	96	40	and	and	CCONJ
bracis-33560	96	41	categorization	categorization	NOUN
bracis-33560	96	42	in	in	ADP
bracis-33560	96	43	content	content	NOUN
bracis-33560	96	44	safety	safety	NOUN
bracis-33560	96	45	pipeline	pipeline	NOUN
bracis-33560	96	46	.	.	PUNCT
bracis-33560	97	1	figure	figure	VERB
bracis-33560	97	2	 	 	SPACE
bracis-33560	97	3	1	1	NUM
bracis-33560	97	4	presents	present	VERB
bracis-33560	97	5	the	the	DET
bracis-33560	97	6	entire	entire	ADJ
bracis-33560	97	7	workflow	workflow	NOUN
bracis-33560	97	8	.	.	PUNCT
bracis-33560	98	1	fig	fig	NOUN
bracis-33560	98	2	.	.	PUNCT
bracis-33560	99	1	1	1	X
bracis-33560	99	2	.	.	X
bracis-33560	99	3	double	double	ADJ
bracis-33560	99	4	pipeline	pipeline	NOUN
bracis-33560	99	5	:	:	PUNCT
bracis-33560	99	6	data	datum	NOUN
bracis-33560	99	7	pipeline	pipeline	NOUN
bracis-33560	99	8	contains	contain	VERB
bracis-33560	99	9	collection	collection	NOUN
bracis-33560	99	10	and	and	CCONJ
bracis-33560	99	11	sampling	sampling	NOUN
bracis-33560	99	12	,	,	PUNCT
bracis-33560	99	13	text	text	NOUN
bracis-33560	99	14	extraction	extraction	NOUN
bracis-33560	99	15	,	,	PUNCT
bracis-33560	99	16	language	language	NOUN
bracis-33560	99	17	identification	identification	NOUN
bracis-33560	99	18	,	,	PUNCT
bracis-33560	99	19	deduplication	deduplication	NOUN
bracis-33560	99	20	,	,	PUNCT
bracis-33560	99	21	and	and	CCONJ
bracis-33560	99	22	quality	quality	NOUN
bracis-33560	99	23	filters	filter	NOUN
bracis-33560	99	24	;	;	PUNCT
bracis-33560	99	25	and	and	CCONJ
bracis-33560	99	26	content	content	NOUN
bracis-33560	99	27	safety	safety	NOUN
bracis-33560	99	28	pipeline	pipeline	NOUN
bracis-33560	99	29	encompasses	encompass	VERB
bracis-33560	99	30	sexual	sexual	ADJ
bracis-33560	99	31	content	content	NOUN
bracis-33560	99	32	filter	filter	NOUN
bracis-33560	99	33	,	,	PUNCT
bracis-33560	99	34	toxic	toxic	ADJ
bracis-33560	99	35	data	datum	NOUN
bracis-33560	99	36	filter	filter	NOUN
bracis-33560	99	37	,	,	PUNCT
bracis-33560	99	38	bias	bias	NOUN
bracis-33560	99	39	filter	filter	NOUN
bracis-33560	99	40	,	,	PUNCT
bracis-33560	99	41	and	and	CCONJ
bracis-33560	99	42	categorization	categorization	NOUN
bracis-33560	99	43	.	.	PUNCT
bracis-33560	100	1	full	full	ADJ
bracis-33560	100	2	size	size	NOUN
bracis-33560	100	3	image	image	NOUN
bracis-33560	100	4	3.1	3.1	NUM
bracis-33560	100	5	data	datum	NOUN
bracis-33560	100	6	collection	collection	NOUN
bracis-33560	100	7	and	and	CCONJ
bracis-33560	100	8	 	 	SPACE
bracis-33560	100	9	sampling	sample	VERB
bracis-33560	100	10	the	the	DET
bracis-33560	100	11	data	datum	NOUN
bracis-33560	100	12	collection	collection	NOUN
bracis-33560	100	13	and	and	CCONJ
bracis-33560	100	14	sampling	sample	VERB
bracis-33560	100	15	step	step	NOUN
bracis-33560	100	16	involves	involve	VERB
bracis-33560	100	17	downloading	download	VERB
bracis-33560	100	18	and	and	CCONJ
bracis-33560	100	19	extracting	extract	VERB
bracis-33560	100	20	portuguese	portuguese	ADJ
bracis-33560	100	21	text	text	NOUN
bracis-33560	100	22	from	from	ADP
bracis-33560	100	23	raw	raw	ADJ
bracis-33560	100	24	web	web	NOUN
bracis-33560	100	25	archive	archive	NOUN
bracis-33560	100	26	(	(	PUNCT
bracis-33560	100	27	warc	warc	NOUN
bracis-33560	100	28	)	)	PUNCT
bracis-33560	100	29	in	in	ADP
bracis-33560	100	30	files	file	NOUN
bracis-33560	100	31	.	.	PUNCT
bracis-33560	101	1	all	all	DET
bracis-33560	101	2	data	datum	NOUN
bracis-33560	101	3	is	be	AUX
bracis-33560	101	4	sourced	source	VERB
bracis-33560	101	5	from	from	ADP
bracis-33560	101	6	common	common	ADJ
bracis-33560	101	7	crawl	crawl	NOUN
bracis-33560	101	8	(	(	PUNCT
bracis-33560	101	9	cc	cc	NOUN
bracis-33560	101	10	)	)	PUNCT
bracis-33560	101	11	,	,	PUNCT
bracis-33560	101	12	which	which	PRON
bracis-33560	101	13	contains	contain	VERB
bracis-33560	101	14	petabytes	petabyte	NOUN
bracis-33560	101	15	of	of	ADP
bracis-33560	101	16	scraped	scrape	VERB
bracis-33560	101	17	internet	internet	NOUN
bracis-33560	101	18	content	content	NOUN
bracis-33560	101	19	from	from	ADP
bracis-33560	101	20	millions	million	NOUN
bracis-33560	101	21	of	of	ADP
bracis-33560	101	22	web	web	NOUN
bracis-33560	101	23	pages	page	NOUN
bracis-33560	101	24	.	.	PUNCT
bracis-33560	102	1	we	we	PRON
bracis-33560	102	2	use	use	VERB
bracis-33560	102	3	the	the	DET
bracis-33560	102	4	raw	raw	ADJ
bracis-33560	102	5	http	http	NOUN
bracis-33560	102	6	files	file	NOUN
bracis-33560	102	7	as	as	ADP
bracis-33560	102	8	the	the	DET
bracis-33560	102	9	initial	initial	ADJ
bracis-33560	102	10	material	material	NOUN
bracis-33560	102	11	,	,	PUNCT
bracis-33560	102	12	from	from	ADP
bracis-33560	102	13	which	which	PRON
bracis-33560	102	14	we	we	PRON
bracis-33560	102	15	extract	extract	VERB
bracis-33560	102	16	and	and	CCONJ
bracis-33560	102	17	filter	filter	NOUN
bracis-33560	102	18	portuguese	portuguese	ADJ
bracis-33560	102	19	text	text	NOUN
bracis-33560	102	20	as	as	SCONJ
bracis-33560	102	21	detailed	detail	VERB
bracis-33560	102	22	in	in	ADP
bracis-33560	102	23	subsects	subsect	NOUN
bracis-33560	102	24	.	.	PUNCT
bracis-33560	102	25	 	 	SPACE
bracis-33560	102	26	3.2	3.2	NUM
bracis-33560	102	27	 	 	SPACE
bracis-33560	102	28	and	and	CCONJ
bracis-33560	102	29	 	 	SPACE
bracis-33560	102	30	3.3	3.3	NUM
bracis-33560	102	31	.	.	PUNCT
bracis-33560	103	1	cc	cc	AUX
bracis-33560	103	2	organizes	organize	VERB
bracis-33560	103	3	its	its	PRON
bracis-33560	103	4	datasets	dataset	NOUN
bracis-33560	103	5	by	by	ADP
bracis-33560	103	6	date	date	NOUN
bracis-33560	103	7	,	,	PUNCT
bracis-33560	103	8	each	each	PRON
bracis-33560	103	9	comprising	comprise	VERB
bracis-33560	103	10	thousands	thousand	NOUN
bracis-33560	103	11	of	of	ADP
bracis-33560	103	12	individual	individual	ADJ
bracis-33560	103	13	shards	shard	NOUN
bracis-33560	103	14	of	of	ADP
bracis-33560	103	15	scraped	scrape	VERB
bracis-33560	103	16	content	content	NOUN
bracis-33560	103	17	.	.	PUNCT
bracis-33560	104	1	we	we	PRON
bracis-33560	104	2	sampled	sample	VERB
bracis-33560	104	3	shards	shard	NOUN
bracis-33560	104	4	from	from	ADP
bracis-33560	104	5	datasets	dataset	NOUN
bracis-33560	104	6	ranging	range	VERB
bracis-33560	104	7	from	from	ADP
bracis-33560	104	8	2015	2015	NUM
bracis-33560	104	9	to	to	ADP
bracis-33560	104	10	2023	2023	NUM
bracis-33560	104	11	,	,	PUNCT
bracis-33560	104	12	prioritizing	prioritize	VERB
bracis-33560	104	13	more	more	ADV
bracis-33560	104	14	recent	recent	ADJ
bracis-33560	104	15	data	datum	NOUN
bracis-33560	104	16	.	.	PUNCT
bracis-33560	105	1	3.2	3.2	NUM
bracis-33560	105	2	text	text	NOUN
bracis-33560	105	3	extraction	extraction	NOUN
bracis-33560	105	4	this	this	DET
bracis-33560	105	5	computational	computational	ADJ
bracis-33560	105	6	step	step	NOUN
bracis-33560	105	7	involves	involve	VERB
bracis-33560	105	8	using	use	VERB
bracis-33560	105	9	multiple	multiple	ADJ
bracis-33560	105	10	cloud	cloud	ADJ
bracis-33560	105	11	machines	machine	NOUN
bracis-33560	105	12	in	in	ADP
bracis-33560	105	13	parallel	parallel	NOUN
bracis-33560	105	14	.	.	PUNCT
bracis-33560	106	1	these	these	DET
bracis-33560	106	2	instances	instance	NOUN
bracis-33560	106	3	download	download	VERB
bracis-33560	106	4	and	and	CCONJ
bracis-33560	106	5	process	process	VERB
bracis-33560	106	6	raw	raw	ADJ
bracis-33560	106	7	files	file	NOUN
bracis-33560	106	8	by	by	ADP
bracis-33560	106	9	extracting	extract	VERB
bracis-33560	106	10	text	text	NOUN
bracis-33560	106	11	from	from	ADP
bracis-33560	106	12	the	the	DET
bracis-33560	106	13	html	html	NOUN
bracis-33560	106	14	and	and	CCONJ
bracis-33560	106	15	filtering	filter	VERB
bracis-33560	106	16	for	for	ADP
bracis-33560	106	17	portuguese	portuguese	ADJ
bracis-33560	106	18	text	text	NOUN
bracis-33560	106	19	.	.	PUNCT
bracis-33560	107	1	the	the	DET
bracis-33560	107	2	resulting	result	VERB
bracis-33560	107	3	data	datum	NOUN
bracis-33560	107	4	is	be	AUX
bracis-33560	107	5	then	then	ADV
bracis-33560	107	6	processed	process	VERB
bracis-33560	107	7	further	far	ADV
bracis-33560	107	8	using	use	VERB
bracis-33560	107	9	a	a	DET
bracis-33560	107	10	single	single	ADJ
bracis-33560	107	11	machine	machine	NOUN
bracis-33560	107	12	containing	contain	VERB
bracis-33560	107	13	a	a	DET
bracis-33560	107	14	key	key	ADJ
bracis-33560	107	15	-	-	PUNCT
bracis-33560	107	16	value	value	NOUN
bracis-33560	107	17	database	database	NOUN
bracis-33560	107	18	for	for	ADP
bracis-33560	107	19	deduplication	deduplication	NOUN
bracis-33560	107	20	,	,	PUNCT
bracis-33560	107	21	as	as	SCONJ
bracis-33560	107	22	explained	explain	VERB
bracis-33560	107	23	in	in	ADP
bracis-33560	107	24	subsect	subsect	NOUN
bracis-33560	107	25	.	.	PUNCT
bracis-33560	107	26	 	 	SPACE
bracis-33560	108	1	3.4	3.4	NUM
bracis-33560	108	2	.	.	PUNCT
bracis-33560	109	1	we	we	PRON
bracis-33560	109	2	opted	opt	VERB
bracis-33560	109	3	to	to	PART
bracis-33560	109	4	work	work	VERB
bracis-33560	109	5	with	with	ADP
bracis-33560	109	6	warc	warc	NOUN
bracis-33560	109	7	files	file	NOUN
bracis-33560	109	8	to	to	PART
bracis-33560	109	9	ensure	ensure	VERB
bracis-33560	109	10	better	well	ADJ
bracis-33560	109	11	quality	quality	NOUN
bracis-33560	109	12	text	text	NOUN
bracis-33560	109	13	,	,	PUNCT
bracis-33560	109	14	which	which	PRON
bracis-33560	109	15	includes	include	VERB
bracis-33560	109	16	handling	handle	VERB
bracis-33560	109	17	raw	raw	ADJ
bracis-33560	109	18	html	html	NOUN
bracis-33560	109	19	files	file	NOUN
bracis-33560	109	20	and	and	CCONJ
bracis-33560	109	21	extracting	extract	VERB
bracis-33560	109	22	texts	text	NOUN
bracis-33560	109	23	ourselves	ourselves	PRON
bracis-33560	109	24	.	.	PUNCT
bracis-33560	110	1	a	a	DET
bracis-33560	110	2	python	python	PROPN
bracis-33560	110	3	library	library	NOUN
bracis-33560	110	4	called	call	VERB
bracis-33560	110	5	trafilatura	trafilatura	ADP
bracis-33560	110	6	 	 	SPACE
bracis-33560	110	7	[	[	X
bracis-33560	110	8	5	5	NUM
bracis-33560	110	9	]	]	PUNCT
bracis-33560	110	10	extracted	extract	VERB
bracis-33560	110	11	only	only	ADV
bracis-33560	110	12	natural	natural	ADJ
bracis-33560	110	13	language	language	NOUN
bracis-33560	110	14	text	text	NOUN
bracis-33560	110	15	from	from	ADP
bracis-33560	110	16	the	the	DET
bracis-33560	110	17	html	html	NOUN
bracis-33560	110	18	files	file	NOUN
bracis-33560	110	19	.	.	PUNCT
bracis-33560	111	1	metadata	metadata	NOUN
bracis-33560	111	2	from	from	ADP
bracis-33560	111	3	webpages	webpage	NOUN
bracis-33560	111	4	were	be	AUX
bracis-33560	111	5	saved	save	VERB
bracis-33560	111	6	for	for	ADP
bracis-33560	111	7	later	later	ADJ
bracis-33560	111	8	use	use	NOUN
bracis-33560	111	9	in	in	ADP
bracis-33560	111	10	the	the	DET
bracis-33560	111	11	pipeline	pipeline	NOUN
bracis-33560	111	12	.	.	PUNCT
bracis-33560	112	1	3.3	3.3	NUM
bracis-33560	112	2	language	language	NOUN
bracis-33560	112	3	identification	identification	NOUN
bracis-33560	112	4	roughly	roughly	ADV
bracis-33560	112	5	0.2	0.2	NUM
bracis-33560	112	6	%	%	NOUN
bracis-33560	112	7	of	of	ADP
bracis-33560	112	8	the	the	DET
bracis-33560	112	9	pages	page	NOUN
bracis-33560	112	10	in	in	ADP
bracis-33560	112	11	each	each	DET
bracis-33560	112	12	shard	shard	NOUN
bracis-33560	112	13	are	be	AUX
bracis-33560	112	14	in	in	ADP
bracis-33560	112	15	portuguese	portuguese	ADJ
bracis-33560	112	16	.	.	PUNCT
bracis-33560	113	1	to	to	PART
bracis-33560	113	2	filter	filter	VERB
bracis-33560	113	3	these	these	DET
bracis-33560	113	4	pages	page	NOUN
bracis-33560	113	5	,	,	PUNCT
bracis-33560	113	6	it	it	PRON
bracis-33560	113	7	is	be	AUX
bracis-33560	113	8	necessary	necessary	ADJ
bracis-33560	113	9	to	to	PART
bracis-33560	113	10	detect	detect	VERB
bracis-33560	113	11	the	the	DET
bracis-33560	113	12	language	language	NOUN
bracis-33560	113	13	they	they	PRON
bracis-33560	113	14	were	be	AUX
bracis-33560	113	15	written	write	VERB
bracis-33560	113	16	in	in	ADP
bracis-33560	113	17	automatically	automatically	ADV
bracis-33560	113	18	.	.	PUNCT
bracis-33560	114	1	as	as	SCONJ
bracis-33560	114	2	utilized	utilize	VERB
bracis-33560	114	3	by	by	ADP
bracis-33560	114	4	 	 	SPACE
bracis-33560	114	5	[	[	X
bracis-33560	114	6	14	14	NUM
bracis-33560	114	7	]	]	PUNCT
bracis-33560	114	8	and	and	CCONJ
bracis-33560	114	9	 	 	SPACE
bracis-33560	115	1	[	[	X
bracis-33560	115	2	1	1	NUM
bracis-33560	115	3	]	]	PUNCT
bracis-33560	115	4	,	,	PUNCT
bracis-33560	115	5	we	we	PRON
bracis-33560	115	6	employ	employ	VERB
bracis-33560	115	7	meta	meta	PROPN
bracis-33560	115	8	ai	ai	VERB
bracis-33560	115	9	’s	’s	PART
bracis-33560	115	10	pre	pre	ADJ
bracis-33560	115	11	-	-	ADJ
bracis-33560	115	12	trained	train	VERB
bracis-33560	115	13	fasttext	fasttext	PROPN
bracis-33560	115	14	model	model	NOUN
bracis-33560	115	15	,	,	PUNCT
bracis-33560	115	16	which	which	PRON
bracis-33560	115	17	can	can	AUX
bracis-33560	115	18	detect	detect	VERB
bracis-33560	115	19	176	176	NUM
bracis-33560	115	20	languages	language	NOUN
bracis-33560	115	21	.	.	PUNCT
bracis-33560	116	1	for	for	ADP
bracis-33560	116	2	each	each	DET
bracis-33560	116	3	downloaded	download	VERB
bracis-33560	116	4	page	page	NOUN
bracis-33560	116	5	,	,	PUNCT
bracis-33560	116	6	the	the	DET
bracis-33560	116	7	text	text	NOUN
bracis-33560	116	8	is	be	AUX
bracis-33560	116	9	extracted	extract	VERB
bracis-33560	116	10	using	use	VERB
bracis-33560	116	11	the	the	DET
bracis-33560	116	12	trafilatura	trafilatura	ADJ
bracis-33560	116	13	library	library	NOUN
bracis-33560	116	14	[	[	X
bracis-33560	116	15	5	5	NUM
bracis-33560	116	16	]	]	PUNCT
bracis-33560	116	17	and	and	CCONJ
bracis-33560	116	18	the	the	DET
bracis-33560	116	19	fasttextfootnote	fasttextfootnote	NOUN
bracis-33560	116	20	2	2	NUM
bracis-33560	116	21	[	[	SYM
bracis-33560	116	22	20	20	NUM
bracis-33560	116	23	]	]	PUNCT
bracis-33560	116	24	models	model	NOUN
bracis-33560	116	25	used	use	VERB
bracis-33560	116	26	to	to	PART
bracis-33560	116	27	determine	determine	VERB
bracis-33560	116	28	the	the	DET
bracis-33560	116	29	language	language	NOUN
bracis-33560	116	30	.	.	PUNCT
bracis-33560	117	1	pages	page	NOUN
bracis-33560	117	2	identified	identify	VERB
bracis-33560	117	3	as	as	ADP
bracis-33560	117	4	portuguese	portuguese	ADJ
bracis-33560	117	5	with	with	ADP
bracis-33560	117	6	the	the	DET
bracis-33560	117	7	highest	high	ADJ
bracis-33560	117	8	probability	probability	NOUN
bracis-33560	117	9	by	by	ADP
bracis-33560	117	10	fasttext	fasttext	PROPN
bracis-33560	117	11	were	be	AUX
bracis-33560	117	12	selected	select	VERB
bracis-33560	117	13	.	.	PUNCT
bracis-33560	118	1	3.4	3.4	NUM
bracis-33560	118	2	deduplication	deduplication	NOUN
bracis-33560	118	3	the	the	DET
bracis-33560	118	4	purpose	purpose	NOUN
bracis-33560	118	5	of	of	ADP
bracis-33560	118	6	this	this	DET
bracis-33560	118	7	step	step	NOUN
bracis-33560	118	8	is	be	AUX
bracis-33560	118	9	to	to	PART
bracis-33560	118	10	remove	remove	VERB
bracis-33560	118	11	duplicated	duplicated	ADJ
bracis-33560	118	12	data	datum	NOUN
bracis-33560	118	13	from	from	ADP
bracis-33560	118	14	the	the	DET
bracis-33560	118	15	corpus	corpus	NOUN
bracis-33560	118	16	.	.	PUNCT
bracis-33560	119	1	to	to	PART
bracis-33560	119	2	achieve	achieve	VERB
bracis-33560	119	3	this	this	DET
bracis-33560	119	4	goal	goal	NOUN
bracis-33560	119	5	,	,	PUNCT
bracis-33560	119	6	we	we	PRON
bracis-33560	119	7	use	use	VERB
bracis-33560	119	8	two	two	NUM
bracis-33560	119	9	deduplication	deduplication	NOUN
bracis-33560	119	10	approaches	approach	NOUN
bracis-33560	119	11	.	.	PUNCT
bracis-33560	120	1	the	the	DET
bracis-33560	120	2	first	first	ADJ
bracis-33560	120	3	is	be	AUX
bracis-33560	120	4	a	a	DET
bracis-33560	120	5	page	page	NOUN
bracis-33560	120	6	-	-	PUNCT
bracis-33560	120	7	level	level	NOUN
bracis-33560	120	8	approach	approach	NOUN
bracis-33560	120	9	,	,	PUNCT
bracis-33560	120	10	which	which	PRON
bracis-33560	120	11	identifies	identify	VERB
bracis-33560	120	12	and	and	CCONJ
bracis-33560	120	13	removes	remove	VERB
bracis-33560	120	14	pages	page	NOUN
bracis-33560	120	15	with	with	ADP
bracis-33560	120	16	duplicate	duplicate	ADJ
bracis-33560	120	17	urls	url	NOUN
bracis-33560	120	18	.	.	PUNCT
bracis-33560	121	1	the	the	DET
bracis-33560	121	2	second	second	NOUN
bracis-33560	121	3	is	be	AUX
bracis-33560	121	4	the	the	DET
bracis-33560	121	5	document	document	NOUN
bracis-33560	121	6	-	-	PUNCT
bracis-33560	121	7	level	level	NOUN
bracis-33560	121	8	approach	approach	NOUN
bracis-33560	121	9	,	,	PUNCT
bracis-33560	121	10	which	which	PRON
bracis-33560	121	11	aims	aim	VERB
bracis-33560	121	12	to	to	PART
bracis-33560	121	13	remove	remove	VERB
bracis-33560	121	14	significant	significant	ADJ
bracis-33560	121	15	overlapping	overlap	VERB
bracis-33560	121	16	documents	document	NOUN
bracis-33560	121	17	.	.	PUNCT
bracis-33560	122	1	we	we	PRON
bracis-33560	122	2	employ	employ	VERB
bracis-33560	122	3	the	the	DET
bracis-33560	122	4	minhashlsh	minhashlsh	NOUN
bracis-33560	122	5	algorithm	algorithm	NOUN
bracis-33560	122	6	to	to	PART
bracis-33560	122	7	calculate	calculate	VERB
bracis-33560	122	8	the	the	DET
bracis-33560	122	9	jaccard	jaccard	ADJ
bracis-33560	122	10	similarity	similarity	NOUN
bracis-33560	122	11	between	between	ADP
bracis-33560	122	12	documents	document	NOUN
bracis-33560	122	13	,	,	PUNCT
bracis-33560	122	14	considering	consider	VERB
bracis-33560	122	15	whether	whether	SCONJ
bracis-33560	122	16	two	two	NUM
bracis-33560	122	17	document	document	NOUN
bracis-33560	122	18	similarity	similarity	NOUN
bracis-33560	122	19	exceeds	exceed	VERB
bracis-33560	122	20	0.7	0.7	NUM
bracis-33560	122	21	 	 	SPACE
bracis-33560	122	22	[	[	X
bracis-33560	122	23	29	29	NUM
bracis-33560	122	24	]	]	PUNCT
bracis-33560	122	25	.	.	PUNCT
bracis-33560	123	1	3.5	3.5	NUM
bracis-33560	123	2	quality	quality	NOUN
bracis-33560	123	3	filters	filter	VERB
bracis-33560	123	4	a	a	DET
bracis-33560	123	5	significant	significant	ADJ
bracis-33560	123	6	amount	amount	NOUN
bracis-33560	123	7	of	of	ADP
bracis-33560	123	8	data	datum	NOUN
bracis-33560	123	9	available	available	ADJ
bracis-33560	123	10	on	on	ADP
bracis-33560	123	11	the	the	DET
bracis-33560	123	12	internet	internet	NOUN
bracis-33560	123	13	may	may	AUX
bracis-33560	123	14	be	be	AUX
bracis-33560	123	15	insufficient	insufficient	ADJ
bracis-33560	123	16	in	in	ADP
bracis-33560	123	17	terms	term	NOUN
bracis-33560	123	18	of	of	ADP
bracis-33560	123	19	quality	quality	NOUN
bracis-33560	123	20	for	for	ADP
bracis-33560	123	21	linguistic	linguistic	ADJ
bracis-33560	123	22	model	model	NOUN
bracis-33560	123	23	formation	formation	NOUN
bracis-33560	123	24	.	.	PUNCT
bracis-33560	124	1	some	some	DET
bracis-33560	124	2	examples	example	NOUN
bracis-33560	124	3	include	include	VERB
bracis-33560	124	4	automatically	automatically	ADV
bracis-33560	124	5	generated	generate	VERB
bracis-33560	124	6	text	text	NOUN
bracis-33560	124	7	and	and	CCONJ
bracis-33560	124	8	text	text	NOUN
bracis-33560	124	9	not	not	PART
bracis-33560	124	10	written	write	VERB
bracis-33560	124	11	for	for	ADP
bracis-33560	124	12	human	human	ADJ
bracis-33560	124	13	consumption	consumption	NOUN
bracis-33560	124	14	 	 	SPACE
bracis-33560	125	1	[	[	X
bracis-33560	125	2	29	29	NUM
bracis-33560	125	3	]	]	PUNCT
bracis-33560	125	4	.	.	PUNCT
bracis-33560	126	1	this	this	DET
bracis-33560	126	2	step	step	NOUN
bracis-33560	126	3	aims	aim	VERB
bracis-33560	126	4	to	to	PART
bracis-33560	126	5	retain	retain	VERB
bracis-33560	126	6	only	only	ADJ
bracis-33560	126	7	pages	page	NOUN
bracis-33560	126	8	written	write	VERB
bracis-33560	126	9	by	by	ADP
bracis-33560	126	10	humans	human	NOUN
bracis-33560	126	11	for	for	ADP
bracis-33560	126	12	humans	human	NOUN
bracis-33560	126	13	.	.	PUNCT
bracis-33560	127	1	to	to	PART
bracis-33560	127	2	achieve	achieve	VERB
bracis-33560	127	3	this	this	PRON
bracis-33560	127	4	,	,	PUNCT
bracis-33560	127	5	we	we	PRON
bracis-33560	127	6	applied	apply	VERB
bracis-33560	127	7	a	a	DET
bracis-33560	127	8	series	series	NOUN
bracis-33560	127	9	of	of	ADP
bracis-33560	127	10	ten	ten	NUM
bracis-33560	127	11	quality	quality	NOUN
bracis-33560	127	12	filters	filter	NOUN
bracis-33560	127	13	:	:	PUNCT
bracis-33560	127	14	number	number	NOUN
bracis-33560	127	15	of	of	ADP
bracis-33560	127	16	tokens	token	NOUN
bracis-33560	127	17	:	:	PUNCT
bracis-33560	127	18	removes	remove	VERB
bracis-33560	127	19	pages	page	NOUN
bracis-33560	127	20	with	with	ADP
bracis-33560	127	21	fewer	few	ADJ
bracis-33560	127	22	than	than	ADP
bracis-33560	127	23	a	a	DET
bracis-33560	127	24	minimum	minimum	ADJ
bracis-33560	127	25	number	number	NOUN
bracis-33560	127	26	of	of	ADP
bracis-33560	127	27	tokens	token	NOUN
bracis-33560	127	28	(	(	PUNCT
bracis-33560	127	29	in	in	ADP
bracis-33560	127	30	this	this	DET
bracis-33560	127	31	work	work	NOUN
bracis-33560	127	32	,	,	PUNCT
bracis-33560	127	33	we	we	PRON
bracis-33560	127	34	used	use	VERB
bracis-33560	127	35	the	the	DET
bracis-33560	127	36	same	same	ADJ
bracis-33560	127	37	tokenizer	tokenizer	NOUN
bracis-33560	127	38	employed	employ	VERB
bracis-33560	127	39	by	by	ADP
bracis-33560	127	40	gpt-2	gpt-2	NOUN
bracis-33560	127	41	)	)	PUNCT
bracis-33560	127	42	,	,	PUNCT
bracis-33560	127	43	as	as	SCONJ
bracis-33560	127	44	texts	text	NOUN
bracis-33560	127	45	with	with	ADP
bracis-33560	127	46	low	low	ADJ
bracis-33560	127	47	token	token	ADJ
bracis-33560	127	48	counts	count	NOUN
bracis-33560	127	49	are	be	AUX
bracis-33560	127	50	generally	generally	ADV
bracis-33560	127	51	not	not	PART
bracis-33560	127	52	informative	informative	ADJ
bracis-33560	127	53	;	;	PUNCT
bracis-33560	127	54	number	number	NOUN
bracis-33560	127	55	of	of	ADP
bracis-33560	127	56	words	word	NOUN
bracis-33560	127	57	:	:	PUNCT
bracis-33560	127	58	removes	remove	VERB
bracis-33560	127	59	pages	page	NOUN
bracis-33560	127	60	that	that	PRON
bracis-33560	127	61	do	do	AUX
bracis-33560	127	62	not	not	PART
bracis-33560	127	63	attend	attend	VERB
bracis-33560	127	64	specified	specify	VERB
bracis-33560	127	65	upper	upper	ADJ
bracis-33560	127	66	and	and	CCONJ
bracis-33560	127	67	lower	low	ADJ
bracis-33560	127	68	word	word	NOUN
bracis-33560	127	69	limits	limit	NOUN
bracis-33560	127	70	,	,	PUNCT
bracis-33560	127	71	excluding	exclude	VERB
bracis-33560	127	72	punctuation	punctuation	NOUN
bracis-33560	127	73	and	and	CCONJ
bracis-33560	127	74	special	special	ADJ
bracis-33560	127	75	characters	character	NOUN
bracis-33560	127	76	;	;	PUNCT
bracis-33560	127	77	type	type	VERB
bracis-33560	127	78	token	token	VERB
bracis-33560	127	79	ratio	ratio	NOUN
bracis-33560	127	80	(	(	PUNCT
bracis-33560	127	81	ttr	ttr	PROPN
bracis-33560	127	82	):	):	PUNCT
bracis-33560	127	83	the	the	DET
bracis-33560	127	84	ratio	ratio	NOUN
bracis-33560	127	85	of	of	ADP
bracis-33560	127	86	unique	unique	ADJ
bracis-33560	127	87	words	word	NOUN
bracis-33560	127	88	(	(	PUNCT
bracis-33560	127	89	types	type	NOUN
bracis-33560	127	90	)	)	PUNCT
bracis-33560	127	91	to	to	PART
bracis-33560	127	92	total	total	ADJ
bracis-33560	127	93	words	word	NOUN
bracis-33560	127	94	(	(	PUNCT
bracis-33560	127	95	tokens	token	NOUN
bracis-33560	127	96	)	)	PUNCT
bracis-33560	127	97	 	 	SPACE
bracis-33560	128	1	[	[	X
bracis-33560	128	2	31	31	NUM
bracis-33560	128	3	]	]	PUNCT
bracis-33560	128	4	.	.	PUNCT
bracis-33560	129	1	ttr	ttr	NOUN
bracis-33560	129	2	 	 	SPACE
bracis-33560	130	1	[	[	X
bracis-33560	130	2	37	37	NUM
bracis-33560	130	3	]	]	PUNCT
bracis-33560	130	4	serves	serve	VERB
bracis-33560	130	5	as	as	ADP
bracis-33560	130	6	an	an	DET
bracis-33560	130	7	indicator	indicator	NOUN
bracis-33560	130	8	of	of	ADP
bracis-33560	130	9	text	text	NOUN
bracis-33560	130	10	quality	quality	NOUN
bracis-33560	130	11	;	;	PUNCT
bracis-33560	130	12	symbols	symbol	NOUN
bracis-33560	130	13	-	-	PUNCT
bracis-33560	130	14	word	word	NOUN
bracis-33560	130	15	ratio	ratio	NOUN
bracis-33560	130	16	:	:	PUNCT
bracis-33560	130	17	removes	remove	VERB
bracis-33560	130	18	pages	page	NOUN
bracis-33560	130	19	whose	whose	DET
bracis-33560	130	20	symbol	symbol	NOUN
bracis-33560	130	21	word	word	NOUN
bracis-33560	130	22	percentages	percentage	NOUN
bracis-33560	130	23	exceed	exceed	VERB
bracis-33560	130	24	limits	limit	NOUN
bracis-33560	130	25	.	.	PUNCT
bracis-33560	131	1	any	any	DET
bracis-33560	131	2	special	special	ADJ
bracis-33560	131	3	character	character	NOUN
bracis-33560	131	4	is	be	AUX
bracis-33560	131	5	considered	consider	VERB
bracis-33560	131	6	a	a	DET
bracis-33560	131	7	symbol	symbol	NOUN
bracis-33560	131	8	;	;	PUNCT
bracis-33560	131	9	symbols	symbol	NOUN
bracis-33560	131	10	at	at	ADP
bracis-33560	131	11	the	the	DET
bracis-33560	131	12	beginning	beginning	NOUN
bracis-33560	131	13	of	of	ADP
bracis-33560	131	14	the	the	DET
bracis-33560	131	15	text	text	NOUN
bracis-33560	131	16	:	:	PUNCT
bracis-33560	131	17	removes	remove	VERB
bracis-33560	131	18	pages	page	NOUN
bracis-33560	131	19	with	with	ADP
bracis-33560	131	20	an	an	DET
bracis-33560	131	21	excessive	excessive	ADJ
bracis-33560	131	22	number	number	NOUN
bracis-33560	131	23	of	of	ADP
bracis-33560	131	24	symbols	symbol	NOUN
bracis-33560	131	25	at	at	ADP
bracis-33560	131	26	the	the	DET
bracis-33560	131	27	beginning	beginning	NOUN
bracis-33560	131	28	of	of	ADP
bracis-33560	131	29	the	the	DET
bracis-33560	131	30	text	text	NOUN
bracis-33560	131	31	;	;	PUNCT
bracis-33560	131	32	stopwords	stopword	NOUN
bracis-33560	131	33	:	:	PUNCT
bracis-33560	131	34	the	the	DET
bracis-33560	131	35	presence	presence	NOUN
bracis-33560	131	36	of	of	ADP
bracis-33560	131	37	stopwords	stopword	NOUN
bracis-33560	131	38	may	may	AUX
bracis-33560	131	39	indicate	indicate	VERB
bracis-33560	131	40	text	text	NOUN
bracis-33560	131	41	coherence	coherence	NOUN
bracis-33560	131	42	 	 	SPACE
bracis-33560	132	1	[	[	X
bracis-33560	132	2	29	29	NUM
bracis-33560	132	3	]	]	SYM
bracis-33560	132	4	;	;	PUNCT
bracis-33560	132	5	n	n	CCONJ
bracis-33560	132	6	-	-	PUNCT
bracis-33560	132	7	gram	gram	NOUN
bracis-33560	132	8	repetition	repetition	NOUN
bracis-33560	132	9	:	:	PUNCT
bracis-33560	132	10	excessive	excessive	ADJ
bracis-33560	132	11	repetition	repetition	NOUN
bracis-33560	132	12	of	of	ADP
bracis-33560	132	13	sentences	sentence	NOUN
bracis-33560	132	14	,	,	PUNCT
bracis-33560	132	15	paragraphs	paragraph	NOUN
bracis-33560	132	16	,	,	PUNCT
bracis-33560	132	17	or	or	CCONJ
bracis-33560	132	18	n	n	CCONJ
bracis-33560	132	19	-	-	PUNCT
bracis-33560	132	20	grams	gram	NOUN
bracis-33560	132	21	indicates	indicate	VERB
bracis-33560	132	22	low	low	ADJ
bracis-33560	132	23	informational	informational	ADJ
bracis-33560	132	24	content	content	NOUN
bracis-33560	132	25	 	 	SPACE
bracis-33560	133	1	[	[	X
bracis-33560	133	2	29	29	NUM
bracis-33560	133	3	]	]	X
bracis-33560	133	4	;	;	PUNCT
bracis-33560	133	5	number	number	NOUN
bracis-33560	133	6	of	of	ADP
bracis-33560	133	7	sentences	sentence	NOUN
bracis-33560	133	8	:	:	PUNCT
bracis-33560	133	9	removes	remove	VERB
bracis-33560	133	10	pages	page	NOUN
bracis-33560	133	11	with	with	ADP
bracis-33560	133	12	fewer	few	ADJ
bracis-33560	133	13	than	than	ADP
bracis-33560	133	14	a	a	DET
bracis-33560	133	15	specified	specified	ADJ
bracis-33560	133	16	number	number	NOUN
bracis-33560	133	17	of	of	ADP
bracis-33560	133	18	sentences	sentence	NOUN
bracis-33560	133	19	;	;	PUNCT
bracis-33560	133	20	lorem	lorem	PROPN
bracis-33560	133	21	ipsum	ipsum	PROPN
bracis-33560	133	22	:	:	PUNCT
bracis-33560	133	23	removes	remove	VERB
bracis-33560	133	24	pages	page	NOUN
bracis-33560	133	25	containing	contain	VERB
bracis-33560	133	26	the	the	DET
bracis-33560	133	27	term	term	NOUN
bracis-33560	133	28	“	"	PUNCT
bracis-33560	133	29	lorem	lorem	X
bracis-33560	133	30	ipsum	ipsum	PROPN
bracis-33560	133	31	”	"	PUNCT
bracis-33560	133	32	 	 	SPACE
bracis-33560	134	1	[	[	X
bracis-33560	134	2	30	30	NUM
bracis-33560	134	3	]	]	PUNCT
bracis-33560	134	4	;	;	PUNCT
bracis-33560	134	5	valid	valid	ADJ
bracis-33560	134	6	words	word	NOUN
bracis-33560	134	7	:	:	PUNCT
bracis-33560	134	8	removes	remove	VERB
bracis-33560	134	9	pages	page	NOUN
bracis-33560	134	10	whose	whose	DET
bracis-33560	134	11	percentage	percentage	NOUN
bracis-33560	134	12	of	of	ADP
bracis-33560	134	13	words	word	NOUN
bracis-33560	134	14	found	find	VERB
bracis-33560	134	15	in	in	ADP
bracis-33560	134	16	a	a	DET
bracis-33560	134	17	language	language	NOUN
bracis-33560	134	18	dictionary	dictionary	NOUN
bracis-33560	134	19	is	be	AUX
bracis-33560	134	20	below	below	ADP
bracis-33560	134	21	a	a	DET
bracis-33560	134	22	specified	specify	VERB
bracis-33560	134	23	threshold	threshold	NOUN
bracis-33560	134	24	.	.	PUNCT
bracis-33560	135	1	the	the	DET
bracis-33560	135	2	thresholds	threshold	NOUN
bracis-33560	135	3	for	for	ADP
bracis-33560	135	4	each	each	DET
bracis-33560	135	5	filter	filter	NOUN
bracis-33560	135	6	are	be	AUX
bracis-33560	135	7	detailed	detail	VERB
bracis-33560	135	8	in	in	ADP
bracis-33560	135	9	sect	sect	NOUN
bracis-33560	135	10	.	.	PUNCT
bracis-33560	135	11	 	 	SPACE
bracis-33560	136	1	4	4	NUM
bracis-33560	136	2	.	.	X
bracis-33560	136	3	3.6	3.6	NUM
bracis-33560	136	4	sexual	sexual	ADJ
bracis-33560	136	5	content	content	NOUN
bracis-33560	136	6	filter	filter	NOUN
bracis-33560	136	7	to	to	PART
bracis-33560	136	8	maintain	maintain	VERB
bracis-33560	136	9	the	the	DET
bracis-33560	136	10	integrity	integrity	NOUN
bracis-33560	136	11	of	of	ADP
bracis-33560	136	12	the	the	DET
bracis-33560	136	13	corpus	corpus	NOUN
bracis-33560	136	14	,	,	PUNCT
bracis-33560	136	15	a	a	DET
bracis-33560	136	16	filter	filter	NOUN
bracis-33560	136	17	was	be	AUX
bracis-33560	136	18	applied	apply	VERB
bracis-33560	136	19	to	to	PART
bracis-33560	136	20	remove	remove	VERB
bracis-33560	136	21	sexual	sexual	ADJ
bracis-33560	136	22	content	content	NOUN
bracis-33560	136	23	from	from	ADP
bracis-33560	136	24	the	the	DET
bracis-33560	136	25	data	datum	NOUN
bracis-33560	136	26	.	.	PUNCT
bracis-33560	137	1	we	we	PRON
bracis-33560	137	2	verified	verify	VERB
bracis-33560	137	3	whether	whether	SCONJ
bracis-33560	137	4	a	a	DET
bracis-33560	137	5	url	url	NOUN
bracis-33560	137	6	was	be	AUX
bracis-33560	137	7	present	present	ADJ
bracis-33560	137	8	in	in	ADP
bracis-33560	137	9	the	the	DET
bracis-33560	137	10	université	université	NOUN
bracis-33560	137	11	toulouse	toulouse	NOUN
bracis-33560	137	12	1footnote	1footnote	NUM
bracis-33560	137	13	3	3	NUM
bracis-33560	137	14	(	(	PUNCT
bracis-33560	137	15	ut1	ut1	NOUN
bracis-33560	137	16	)	)	PUNCT
bracis-33560	137	17	blocklist	blocklist	NOUN
bracis-33560	137	18	for	for	ADP
bracis-33560	137	19	each	each	DET
bracis-33560	137	20	page	page	NOUN
bracis-33560	137	21	collected	collect	VERB
bracis-33560	137	22	.	.	PUNCT
bracis-33560	138	1	as	as	SCONJ
bracis-33560	138	2	noted	note	VERB
bracis-33560	138	3	by	by	ADP
bracis-33560	138	4	 	 	SPACE
bracis-33560	138	5	[	[	X
bracis-33560	138	6	2	2	NUM
bracis-33560	138	7	]	]	PUNCT
bracis-33560	138	8	,	,	PUNCT
bracis-33560	138	9	the	the	DET
bracis-33560	138	10	ut1	ut1	NOUN
bracis-33560	138	11	blocklist	blocklist	NOUN
bracis-33560	138	12	is	be	AUX
bracis-33560	138	13	an	an	DET
bracis-33560	138	14	extensive	extensive	ADJ
bracis-33560	138	15	compilation	compilation	NOUN
bracis-33560	138	16	of	of	ADP
bracis-33560	138	17	block	block	NOUN
bracis-33560	138	18	lists	list	NOUN
bracis-33560	138	19	frequently	frequently	ADV
bracis-33560	138	20	used	use	VERB
bracis-33560	138	21	for	for	ADP
bracis-33560	138	22	internet	internet	NOUN
bracis-33560	138	23	access	access	NOUN
bracis-33560	138	24	control	control	NOUN
bracis-33560	138	25	at	at	ADP
bracis-33560	138	26	schools	school	NOUN
bracis-33560	138	27	.	.	PUNCT
bracis-33560	139	1	it	it	PRON
bracis-33560	139	2	was	be	AUX
bracis-33560	139	3	developed	develop	VERB
bracis-33560	139	4	with	with	ADP
bracis-33560	139	5	the	the	DET
bracis-33560	139	6	help	help	NOUN
bracis-33560	139	7	of	of	ADP
bracis-33560	139	8	automated	automate	VERB
bracis-33560	139	9	systems	system	NOUN
bracis-33560	139	10	and	and	CCONJ
bracis-33560	139	11	human	human	ADJ
bracis-33560	139	12	contributors	contributor	NOUN
bracis-33560	139	13	and	and	CCONJ
bracis-33560	139	14	currently	currently	ADV
bracis-33560	139	15	includes	include	VERB
bracis-33560	139	16	3.7	3.7	NUM
bracis-33560	139	17	million	million	NUM
bracis-33560	139	18	entries	entry	NOUN
bracis-33560	139	19	.	.	PUNCT
bracis-33560	140	1	for	for	ADP
bracis-33560	140	2	this	this	DET
bracis-33560	140	3	work	work	NOUN
bracis-33560	140	4	,	,	PUNCT
bracis-33560	140	5	we	we	PRON
bracis-33560	140	6	utilized	utilize	VERB
bracis-33560	140	7	a	a	DET
bracis-33560	140	8	filtered	filter	VERB
bracis-33560	140	9	version	version	NOUN
bracis-33560	140	10	of	of	ADP
bracis-33560	140	11	this	this	DET
bracis-33560	140	12	blocklist	blocklist	NOUN
bracis-33560	140	13	tailored	tailor	VERB
bracis-33560	140	14	for	for	ADP
bracis-33560	140	15	brazilian	brazilian	ADJ
bracis-33560	140	16	websites	website	NOUN
bracis-33560	140	17	.	.	PUNCT
bracis-33560	141	1	it	it	PRON
bracis-33560	141	2	should	should	AUX
bracis-33560	141	3	be	be	AUX
bracis-33560	141	4	noted	note	VERB
bracis-33560	141	5	that	that	SCONJ
bracis-33560	141	6	this	this	DET
bracis-33560	141	7	filter	filter	NOUN
bracis-33560	141	8	only	only	ADV
bracis-33560	141	9	excludes	exclude	VERB
bracis-33560	141	10	websites	website	NOUN
bracis-33560	141	11	marked	mark	VERB
bracis-33560	141	12	as	as	ADP
bracis-33560	141	13	adult	adult	NOUN
bracis-33560	141	14	content	content	NOUN
bracis-33560	141	15	.	.	PUNCT
bracis-33560	142	1	for	for	ADP
bracis-33560	142	2	the	the	DET
bracis-33560	142	3	remaining	remain	VERB
bracis-33560	142	4	content	content	NOUN
bracis-33560	142	5	,	,	PUNCT
bracis-33560	142	6	we	we	PRON
bracis-33560	142	7	randomly	randomly	ADV
bracis-33560	142	8	selected	select	VERB
bracis-33560	142	9	25,000	25,000	NUM
bracis-33560	142	10	examples	example	NOUN
bracis-33560	142	11	and	and	CCONJ
bracis-33560	142	12	used	use	VERB
bracis-33560	142	13	a	a	DET
bracis-33560	142	14	mistral	mistral	ADJ
bracis-33560	142	15	7b	7b	NOUN
bracis-33560	142	16	 	 	SPACE
bracis-33560	142	17	[	[	X
bracis-33560	142	18	19	19	NUM
bracis-33560	142	19	]	]	PUNCT
bracis-33560	142	20	model	model	NOUN
bracis-33560	142	21	to	to	PART
bracis-33560	142	22	extract	extract	VERB
bracis-33560	142	23	pejorative	pejorative	ADJ
bracis-33560	142	24	sexual	sexual	ADJ
bracis-33560	142	25	terms	term	NOUN
bracis-33560	142	26	.	.	PUNCT
bracis-33560	143	1	these	these	DET
bracis-33560	143	2	terms	term	NOUN
bracis-33560	143	3	were	be	AUX
bracis-33560	143	4	then	then	ADV
bracis-33560	143	5	reviewed	review	VERB
bracis-33560	143	6	by	by	ADP
bracis-33560	143	7	humans	human	NOUN
bracis-33560	143	8	and	and	CCONJ
bracis-33560	143	9	used	use	VERB
bracis-33560	143	10	as	as	ADP
bracis-33560	143	11	the	the	DET
bracis-33560	143	12	final	final	ADJ
bracis-33560	143	13	sexual	sexual	ADJ
bracis-33560	143	14	content	content	NOUN
bracis-33560	143	15	filter	filter	NOUN
bracis-33560	143	16	.	.	PUNCT
bracis-33560	144	1	3.7	3.7	NUM
bracis-33560	144	2	toxic	toxic	ADJ
bracis-33560	144	3	data	datum	NOUN
bracis-33560	144	4	filter	filter	NOUN
bracis-33560	144	5	in	in	ADP
bracis-33560	144	6	this	this	DET
bracis-33560	144	7	step	step	NOUN
bracis-33560	144	8	,	,	PUNCT
bracis-33560	144	9	our	our	PRON
bracis-33560	144	10	objective	objective	NOUN
bracis-33560	144	11	is	be	AUX
bracis-33560	144	12	to	to	PART
bracis-33560	144	13	identify	identify	VERB
bracis-33560	144	14	and	and	CCONJ
bracis-33560	144	15	remove	remove	VERB
bracis-33560	144	16	potential	potential	ADJ
bracis-33560	144	17	toxic	toxic	ADJ
bracis-33560	144	18	content	content	NOUN
bracis-33560	144	19	.	.	PUNCT
bracis-33560	145	1	toxicity	toxicity	NOUN
bracis-33560	145	2	definition	definition	NOUN
bracis-33560	145	3	is	be	AUX
bracis-33560	145	4	a	a	DET
bracis-33560	145	5	rude	rude	ADJ
bracis-33560	145	6	,	,	PUNCT
bracis-33560	145	7	disrespectful	disrespectful	ADJ
bracis-33560	145	8	,	,	PUNCT
bracis-33560	145	9	or	or	CCONJ
bracis-33560	145	10	unreasonable	unreasonable	ADJ
bracis-33560	145	11	comment	comment	NOUN
bracis-33560	145	12	likely	likely	ADJ
bracis-33560	145	13	to	to	PART
bracis-33560	145	14	incite	incite	VERB
bracis-33560	145	15	an	an	DET
bracis-33560	145	16	argument	argument	NOUN
bracis-33560	145	17	 	 	SPACE
bracis-33560	146	1	[	[	X
bracis-33560	146	2	13	13	NUM
bracis-33560	146	3	]	]	PUNCT
bracis-33560	146	4	.	.	PUNCT
bracis-33560	147	1	our	our	PRON
bracis-33560	147	2	filter	filter	NOUN
bracis-33560	147	3	comprises	comprise	VERB
bracis-33560	147	4	a	a	DET
bracis-33560	147	5	dictionary	dictionary	NOUN
bracis-33560	147	6	of	of	ADP
bracis-33560	147	7	insults	insult	NOUN
bracis-33560	147	8	and	and	CCONJ
bracis-33560	147	9	pejorative	pejorative	ADJ
bracis-33560	147	10	terms	term	NOUN
bracis-33560	147	11	.	.	PUNCT
bracis-33560	148	1	we	we	PRON
bracis-33560	148	2	evaluated	evaluate	VERB
bracis-33560	148	3	exact	exact	ADJ
bracis-33560	148	4	matches	match	NOUN
bracis-33560	148	5	of	of	ADP
bracis-33560	148	6	dictionary	dictionary	ADJ
bracis-33560	148	7	words	word	NOUN
bracis-33560	148	8	with	with	ADP
bracis-33560	148	9	document	document	NOUN
bracis-33560	148	10	terms	term	NOUN
bracis-33560	148	11	and	and	CCONJ
bracis-33560	148	12	removed	remove	VERB
bracis-33560	148	13	documents	document	NOUN
bracis-33560	148	14	that	that	PRON
bracis-33560	148	15	exceeded	exceed	VERB
bracis-33560	148	16	a	a	DET
bracis-33560	148	17	specified	specify	VERB
bracis-33560	148	18	threshold	threshold	NOUN
bracis-33560	148	19	percentage	percentage	NOUN
bracis-33560	148	20	of	of	ADP
bracis-33560	148	21	words	word	NOUN
bracis-33560	148	22	in	in	ADP
bracis-33560	148	23	the	the	DET
bracis-33560	148	24	dictionary	dictionary	NOUN
bracis-33560	148	25	.	.	PUNCT
bracis-33560	149	1	the	the	DET
bracis-33560	149	2	dictionary	dictionary	PROPN
bracis-33560	149	3	used	use	VERB
bracis-33560	149	4	for	for	ADP
bracis-33560	149	5	this	this	DET
bracis-33560	149	6	filter	filter	NOUN
bracis-33560	149	7	was	be	AUX
bracis-33560	149	8	created	create	VERB
bracis-33560	149	9	by	by	ADP
bracis-33560	149	10	merging	merge	VERB
bracis-33560	149	11	two	two	NUM
bracis-33560	149	12	lists	list	NOUN
bracis-33560	149	13	of	of	ADP
bracis-33560	149	14	wordsfootnote	wordsfootnote	ADJ
bracis-33560	149	15	4\(^,\)footnote	4\(^,\)footnote	NOUN
bracis-33560	149	16	5	5	NUM
bracis-33560	149	17	.	.	PUNCT
bracis-33560	150	1	we	we	PRON
bracis-33560	150	2	reinforce	reinforce	VERB
bracis-33560	150	3	this	this	DET
bracis-33560	150	4	filter	filter	NOUN
bracis-33560	150	5	does	do	AUX
bracis-33560	150	6	not	not	PART
bracis-33560	150	7	aim	aim	VERB
bracis-33560	150	8	to	to	PART
bracis-33560	150	9	eliminate	eliminate	VERB
bracis-33560	150	10	all	all	DET
bracis-33560	150	11	the	the	DET
bracis-33560	150	12	data	datum	NOUN
bracis-33560	150	13	containing	contain	VERB
bracis-33560	150	14	toxic	toxic	ADJ
bracis-33560	150	15	words	word	NOUN
bracis-33560	150	16	,	,	PUNCT
bracis-33560	150	17	but	but	CCONJ
bracis-33560	150	18	rather	rather	ADV
bracis-33560	150	19	to	to	PART
bracis-33560	150	20	remove	remove	VERB
bracis-33560	150	21	content	content	NOUN
bracis-33560	150	22	with	with	ADP
bracis-33560	150	23	a	a	DET
bracis-33560	150	24	significant	significant	ADJ
bracis-33560	150	25	toxic	toxic	ADJ
bracis-33560	150	26	content	content	NOUN
bracis-33560	150	27	proportion	proportion	NOUN
bracis-33560	150	28	.	.	PUNCT
bracis-33560	151	1	3.8	3.8	NUM
bracis-33560	151	2	bias	bias	NOUN
bracis-33560	151	3	filter	filter	VERB
bracis-33560	151	4	our	our	PRON
bracis-33560	151	5	work	work	NOUN
bracis-33560	151	6	involves	involve	VERB
bracis-33560	151	7	a	a	DET
bracis-33560	151	8	step	step	NOUN
bracis-33560	151	9	to	to	PART
bracis-33560	151	10	identify	identify	VERB
bracis-33560	151	11	and	and	CCONJ
bracis-33560	151	12	eliminate	eliminate	VERB
bracis-33560	151	13	potential	potential	ADJ
bracis-33560	151	14	biases	bias	NOUN
bracis-33560	151	15	in	in	ADP
bracis-33560	151	16	the	the	DET
bracis-33560	151	17	text	text	NOUN
bracis-33560	151	18	based	base	VERB
bracis-33560	151	19	on	on	ADP
bracis-33560	151	20	contextual	contextual	ADJ
bracis-33560	151	21	cues	cue	NOUN
bracis-33560	151	22	.	.	PUNCT
bracis-33560	152	1	we	we	PRON
bracis-33560	152	2	construct	construct	VERB
bracis-33560	152	3	a	a	DET
bracis-33560	152	4	dictionary	dictionary	NOUN
bracis-33560	152	5	of	of	ADP
bracis-33560	152	6	portuguese	portuguese	ADJ
bracis-33560	152	7	expressions	expression	NOUN
bracis-33560	152	8	used	use	VERB
bracis-33560	152	9	in	in	ADP
bracis-33560	152	10	biased	biased	ADJ
bracis-33560	152	11	contexts	contexts	NOUN
bracis-33560	152	12	.	.	PUNCT
bracis-33560	153	1	when	when	SCONJ
bracis-33560	153	2	compiling	compile	VERB
bracis-33560	153	3	this	this	DET
bracis-33560	153	4	dictionary	dictionary	NOUN
bracis-33560	153	5	,	,	PUNCT
bracis-33560	153	6	it	it	PRON
bracis-33560	153	7	is	be	AUX
bracis-33560	153	8	crucial	crucial	ADJ
bracis-33560	153	9	to	to	PART
bracis-33560	153	10	consider	consider	VERB
bracis-33560	153	11	the	the	DET
bracis-33560	153	12	society	society	NOUN
bracis-33560	153	13	dynamic	dynamic	ADJ
bracis-33560	153	14	and	and	CCONJ
bracis-33560	153	15	its	its	PRON
bracis-33560	153	16	relationship	relationship	NOUN
bracis-33560	153	17	with	with	ADP
bracis-33560	153	18	linguistics	linguistic	NOUN
bracis-33560	153	19	[	[	X
bracis-33560	153	20	6	6	NUM
bracis-33560	153	21	]	]	PUNCT
bracis-33560	153	22	.	.	PUNCT
bracis-33560	154	1	we	we	PRON
bracis-33560	154	2	filtered	filter	VERB
bracis-33560	154	3	the	the	DET
bracis-33560	154	4	corpus	corpus	NOUN
bracis-33560	154	5	by	by	ADP
bracis-33560	154	6	checking	check	VERB
bracis-33560	154	7	for	for	ADP
bracis-33560	154	8	exact	exact	ADJ
bracis-33560	154	9	matches	match	NOUN
bracis-33560	154	10	between	between	ADP
bracis-33560	154	11	the	the	DET
bracis-33560	154	12	dictionary	dictionary	ADJ
bracis-33560	154	13	words	word	NOUN
bracis-33560	154	14	and	and	CCONJ
bracis-33560	154	15	those	those	PRON
bracis-33560	154	16	in	in	ADP
bracis-33560	154	17	the	the	DET
bracis-33560	154	18	text	text	NOUN
bracis-33560	154	19	 	 	SPACE
bracis-33560	155	1	[	[	X
bracis-33560	155	2	16	16	NUM
bracis-33560	155	3	]	]	PUNCT
bracis-33560	155	4	.	.	PUNCT
bracis-33560	156	1	various	various	ADJ
bracis-33560	156	2	types	type	NOUN
bracis-33560	156	3	of	of	ADP
bracis-33560	156	4	social	social	ADJ
bracis-33560	156	5	biases	bias	NOUN
bracis-33560	156	6	were	be	AUX
bracis-33560	156	7	mapped	map	VERB
bracis-33560	156	8	,	,	PUNCT
bracis-33560	156	9	including	include	VERB
bracis-33560	156	10	gender	gender	NOUN
bracis-33560	156	11	,	,	PUNCT
bracis-33560	156	12	religion	religion	NOUN
bracis-33560	156	13	,	,	PUNCT
bracis-33560	156	14	race	race	NOUN
bracis-33560	156	15	,	,	PUNCT
bracis-33560	156	16	sexist	sexist	ADJ
bracis-33560	156	17	expressions	expression	NOUN
bracis-33560	156	18	,	,	PUNCT
bracis-33560	156	19	xenophobia	xenophobia	NOUN
bracis-33560	156	20	,	,	PUNCT
bracis-33560	156	21	homophobia	homophobia	NOUN
bracis-33560	156	22	,	,	PUNCT
bracis-33560	156	23	ableism	ableism	NOUN
bracis-33560	156	24	,	,	PUNCT
bracis-33560	156	25	fatphobia	fatphobia	PROPN
bracis-33560	156	26	,	,	PUNCT
bracis-33560	156	27	and	and	CCONJ
bracis-33560	156	28	politics	politic	NOUN
bracis-33560	156	29	 	 	SPACE
bracis-33560	157	1	[	[	X
bracis-33560	157	2	24	24	NUM
bracis-33560	157	3	,	,	PUNCT
bracis-33560	157	4	26	26	NUM
bracis-33560	157	5	]	]	PUNCT
bracis-33560	157	6	.	.	PUNCT
bracis-33560	158	1	3.9	3.9	NUM
bracis-33560	158	2	categorization	categorization	NOUN
bracis-33560	158	3	this	this	DET
bracis-33560	158	4	phase	phase	NOUN
bracis-33560	158	5	aims	aim	VERB
bracis-33560	158	6	to	to	PART
bracis-33560	158	7	categorize	categorize	VERB
bracis-33560	158	8	each	each	DET
bracis-33560	158	9	page	page	NOUN
bracis-33560	158	10	into	into	ADP
bracis-33560	158	11	a	a	DET
bracis-33560	158	12	specific	specific	ADJ
bracis-33560	158	13	knowledge	knowledge	NOUN
bracis-33560	158	14	domain	domain	NOUN
bracis-33560	158	15	.	.	PUNCT
bracis-33560	159	1	we	we	PRON
bracis-33560	159	2	identified	identify	VERB
bracis-33560	159	3	27	27	NUM
bracis-33560	159	4	knowledge	knowledge	NOUN
bracis-33560	159	5	domains	domain	NOUN
bracis-33560	159	6	,	,	PUNCT
bracis-33560	159	7	covering	cover	VERB
bracis-33560	159	8	several	several	ADJ
bracis-33560	159	9	categories	category	NOUN
bracis-33560	159	10	and	and	CCONJ
bracis-33560	159	11	subjects	subject	NOUN
bracis-33560	159	12	,	,	PUNCT
bracis-33560	159	13	such	such	ADJ
bracis-33560	159	14	as	as	ADP
bracis-33560	159	15	:	:	PUNCT
bracis-33560	159	16	blog	blog	NOUN
bracis-33560	159	17	posts	post	NOUN
bracis-33560	159	18	,	,	PUNCT
bracis-33560	159	19	news	news	NOUN
bracis-33560	159	20	articles	article	NOUN
bracis-33560	159	21	,	,	PUNCT
bracis-33560	159	22	marketing	marketing	NOUN
bracis-33560	159	23	,	,	PUNCT
bracis-33560	159	24	movies	movie	NOUN
bracis-33560	159	25	,	,	PUNCT
bracis-33560	159	26	social	social	ADJ
bracis-33560	159	27	media	medium	NOUN
bracis-33560	159	28	,	,	PUNCT
bracis-33560	159	29	health	health	NOUN
bracis-33560	159	30	,	,	PUNCT
bracis-33560	159	31	culinary	culinary	ADJ
bracis-33560	159	32	recipes	recipe	NOUN
bracis-33560	159	33	,	,	PUNCT
bracis-33560	159	34	books	book	NOUN
bracis-33560	159	35	,	,	PUNCT
bracis-33560	159	36	scientific	scientific	ADJ
bracis-33560	159	37	articles	article	NOUN
bracis-33560	159	38	,	,	PUNCT
bracis-33560	159	39	politics	politic	NOUN
bracis-33560	159	40	,	,	PUNCT
bracis-33560	159	41	etc	etc	X
bracis-33560	159	42	.	.	X
bracis-33560	160	1	the	the	DET
bracis-33560	160	2	information	information	NOUN
bracis-33560	160	3	regarding	regard	VERB
bracis-33560	160	4	each	each	DET
bracis-33560	160	5	domain	domain	NOUN
bracis-33560	160	6	can	can	AUX
bracis-33560	160	7	be	be	AUX
bracis-33560	160	8	leveraged	leverage	VERB
bracis-33560	160	9	to	to	PART
bracis-33560	160	10	balance	balance	VERB
bracis-33560	160	11	the	the	DET
bracis-33560	160	12	data	datum	NOUN
bracis-33560	160	13	for	for	ADP
bracis-33560	160	14	specific	specific	ADJ
bracis-33560	160	15	tasks	task	NOUN
bracis-33560	160	16	or	or	CCONJ
bracis-33560	160	17	augment	augment	VERB
bracis-33560	160	18	datasets	dataset	NOUN
bracis-33560	160	19	where	where	SCONJ
bracis-33560	160	20	knowledge	knowledge	NOUN
bracis-33560	160	21	is	be	AUX
bracis-33560	160	22	lacking	lack	VERB
bracis-33560	160	23	.	.	PUNCT
bracis-33560	161	1	essentially	essentially	ADV
bracis-33560	161	2	,	,	PUNCT
bracis-33560	161	3	we	we	PRON
bracis-33560	161	4	map	map	VERB
bracis-33560	161	5	the	the	DET
bracis-33560	161	6	urls	url	NOUN
bracis-33560	161	7	of	of	ADP
bracis-33560	161	8	different	different	ADJ
bracis-33560	161	9	pages	page	NOUN
bracis-33560	161	10	to	to	ADP
bracis-33560	161	11	each	each	DET
bracis-33560	161	12	knowledge	knowledge	NOUN
bracis-33560	161	13	scope	scope	NOUN
bracis-33560	161	14	and	and	CCONJ
bracis-33560	161	15	assign	assign	VERB
bracis-33560	161	16	a	a	DET
bracis-33560	161	17	topic	topic	NOUN
bracis-33560	161	18	to	to	ADP
bracis-33560	161	19	each	each	DET
bracis-33560	161	20	url	url	NOUN
bracis-33560	161	21	.	.	PUNCT
bracis-33560	162	1	a	a	DET
bracis-33560	162	2	pt	pt	PROPN
bracis-33560	162	3	-	-	PUNCT
bracis-33560	162	4	pt	pt	NOUN
bracis-33560	162	5	text	text	NOUN
bracis-33560	162	6	category	category	NOUN
bracis-33560	162	7	has	have	AUX
bracis-33560	162	8	been	be	AUX
bracis-33560	162	9	introduced	introduce	VERB
bracis-33560	162	10	to	to	PART
bracis-33560	162	11	differentiate	differentiate	VERB
bracis-33560	162	12	between	between	ADP
bracis-33560	162	13	brazilian	brazilian	ADJ
bracis-33560	162	14	portuguese	portuguese	NOUN
bracis-33560	162	15	and	and	CCONJ
bracis-33560	162	16	european	european	PROPN
bracis-33560	162	17	portuguese	portuguese	PROPN
bracis-33560	162	18	,	,	PUNCT
bracis-33560	162	19	as	as	SCONJ
bracis-33560	162	20	brazilian	brazilian	ADJ
bracis-33560	162	21	portuguese	portuguese	NOUN
bracis-33560	162	22	predominates	predominate	VERB
bracis-33560	162	23	in	in	ADP
bracis-33560	162	24	the	the	DET
bracis-33560	162	25	dataset	dataset	NOUN
bracis-33560	162	26	.	.	PUNCT
bracis-33560	163	1	the	the	DET
bracis-33560	163	2	identified	identify	VERB
bracis-33560	163	3	knowledge	knowledge	NOUN
bracis-33560	163	4	domains	domain	NOUN
bracis-33560	163	5	and	and	CCONJ
bracis-33560	163	6	their	their	PRON
bracis-33560	163	7	distribution	distribution	NOUN
bracis-33560	163	8	are	be	AUX
bracis-33560	163	9	discussed	discuss	VERB
bracis-33560	163	10	in	in	ADP
bracis-33560	163	11	the	the	DET
bracis-33560	163	12	subsect	subsect	NOUN
bracis-33560	163	13	.	.	PUNCT
bracis-33560	163	14	 	 	SPACE
bracis-33560	164	1	5.2	5.2	NUM
bracis-33560	164	2	.	.	PUNCT
bracis-33560	164	3	4	4	NUM
bracis-33560	164	4	qualitative	qualitative	ADJ
bracis-33560	164	5	configuration	configuration	NOUN
bracis-33560	164	6	test	test	NOUN
bracis-33560	164	7	we	we	PRON
bracis-33560	164	8	generated	generate	VERB
bracis-33560	164	9	a	a	DET
bracis-33560	164	10	1	1	NUM
bracis-33560	164	11	gb	gb	NOUN
bracis-33560	164	12	sample	sample	NOUN
bracis-33560	164	13	of	of	ADP
bracis-33560	164	14	texts	text	NOUN
bracis-33560	164	15	and	and	CCONJ
bracis-33560	164	16	analyzed	analyze	VERB
bracis-33560	164	17	the	the	DET
bracis-33560	164	18	distribution	distribution	NOUN
bracis-33560	164	19	of	of	ADP
bracis-33560	164	20	metrics	metric	NOUN
bracis-33560	164	21	such	such	ADJ
bracis-33560	164	22	as	as	ADP
bracis-33560	164	23	the	the	DET
bracis-33560	164	24	number	number	NOUN
bracis-33560	164	25	of	of	ADP
bracis-33560	164	26	tokens	token	NOUN
bracis-33560	164	27	,	,	PUNCT
bracis-33560	164	28	word	word	NOUN
bracis-33560	164	29	count	count	NOUN
bracis-33560	164	30	,	,	PUNCT
bracis-33560	164	31	and	and	CCONJ
bracis-33560	164	32	ttr	ttr	NOUN
bracis-33560	164	33	.	.	PUNCT
bracis-33560	165	1	we	we	PRON
bracis-33560	165	2	use	use	VERB
bracis-33560	165	3	this	this	DET
bracis-33560	165	4	sample	sample	NOUN
bracis-33560	165	5	to	to	PART
bracis-33560	165	6	find	find	VERB
bracis-33560	165	7	the	the	DET
bracis-33560	165	8	best	good	ADJ
bracis-33560	165	9	configuration	configuration	NOUN
bracis-33560	165	10	for	for	ADP
bracis-33560	165	11	our	our	PRON
bracis-33560	165	12	double	double	ADJ
bracis-33560	165	13	-	-	PUNCT
bracis-33560	165	14	pipeline	pipeline	NOUN
bracis-33560	165	15	to	to	PART
bracis-33560	165	16	produce	produce	VERB
bracis-33560	165	17	the	the	DET
bracis-33560	165	18	resultant	resultant	NOUN
bracis-33560	165	19	datasets	dataset	NOUN
bracis-33560	165	20	.	.	PUNCT
bracis-33560	166	1	different	different	ADJ
bracis-33560	166	2	value	value	NOUN
bracis-33560	166	3	sets	set	NOUN
bracis-33560	166	4	were	be	AUX
bracis-33560	166	5	empirically	empirically	ADV
bracis-33560	166	6	tested	test	VERB
bracis-33560	166	7	,	,	PUNCT
bracis-33560	166	8	and	and	CCONJ
bracis-33560	166	9	for	for	ADP
bracis-33560	166	10	each	each	DET
bracis-33560	166	11	one	one	NOUN
bracis-33560	166	12	,	,	PUNCT
bracis-33560	166	13	we	we	PRON
bracis-33560	166	14	checked	check	VERB
bracis-33560	166	15	the	the	DET
bracis-33560	166	16	correctness	correctness	NOUN
bracis-33560	166	17	of	of	ADP
bracis-33560	166	18	page	page	NOUN
bracis-33560	166	19	removals	removal	NOUN
bracis-33560	166	20	and	and	CCONJ
bracis-33560	166	21	recorded	record	VERB
bracis-33560	166	22	the	the	DET
bracis-33560	166	23	number	number	NOUN
bracis-33560	166	24	of	of	ADP
bracis-33560	166	25	pages	page	NOUN
bracis-33560	166	26	excluded	exclude	VERB
bracis-33560	166	27	after	after	SCONJ
bracis-33560	166	28	the	the	DET
bracis-33560	166	29	filters	filter	NOUN
bracis-33560	166	30	were	be	AUX
bracis-33560	166	31	applied	apply	VERB
bracis-33560	166	32	.	.	PUNCT
bracis-33560	167	1	analysis	analysis	NOUN
bracis-33560	167	2	was	be	AUX
bracis-33560	167	3	carried	carry	VERB
bracis-33560	167	4	out	out	ADP
bracis-33560	167	5	on	on	ADP
bracis-33560	167	6	this	this	DET
bracis-33560	167	7	sample	sample	NOUN
bracis-33560	167	8	to	to	PART
bracis-33560	167	9	find	find	VERB
bracis-33560	167	10	the	the	DET
bracis-33560	167	11	optimal	optimal	ADJ
bracis-33560	167	12	configuration	configuration	NOUN
bracis-33560	167	13	.	.	PUNCT
bracis-33560	168	1	moreover	moreover	ADV
bracis-33560	168	2	,	,	PUNCT
bracis-33560	168	3	we	we	PRON
bracis-33560	168	4	perform	perform	VERB
bracis-33560	168	5	qualitative	qualitative	ADJ
bracis-33560	168	6	analyses	analysis	NOUN
bracis-33560	168	7	to	to	PART
bracis-33560	168	8	verify	verify	VERB
bracis-33560	168	9	the	the	DET
bracis-33560	168	10	removal	removal	NOUN
bracis-33560	168	11	appropriateness	appropriateness	NOUN
bracis-33560	168	12	.	.	PUNCT
bracis-33560	169	1	this	this	DET
bracis-33560	169	2	qualitative	qualitative	ADJ
bracis-33560	169	3	test	test	NOUN
bracis-33560	169	4	helped	help	VERB
bracis-33560	169	5	us	we	PRON
bracis-33560	169	6	define	define	VERB
bracis-33560	169	7	satisfactory	satisfactory	ADJ
bracis-33560	169	8	values	value	NOUN
bracis-33560	169	9	that	that	PRON
bracis-33560	169	10	we	we	PRON
bracis-33560	169	11	should	should	AUX
bracis-33560	169	12	filter	filter	VERB
bracis-33560	169	13	to	to	PART
bracis-33560	169	14	obtain	obtain	VERB
bracis-33560	169	15	content	content	NOUN
bracis-33560	169	16	with	with	ADP
bracis-33560	169	17	good	good	ADJ
bracis-33560	169	18	textual	textual	ADJ
bracis-33560	169	19	quality	quality	NOUN
bracis-33560	169	20	,	,	PUNCT
bracis-33560	169	21	i.e.	i.e.	X
bracis-33560	169	22	,	,	PUNCT
bracis-33560	169	23	a	a	DET
bracis-33560	169	24	text	text	NOUN
bracis-33560	169	25	that	that	PRON
bracis-33560	169	26	is	be	AUX
bracis-33560	169	27	diverse	diverse	ADJ
bracis-33560	169	28	in	in	ADP
bracis-33560	169	29	words	word	NOUN
bracis-33560	169	30	,	,	PUNCT
bracis-33560	169	31	fluid	fluid	NOUN
bracis-33560	169	32	,	,	PUNCT
bracis-33560	169	33	with	with	ADP
bracis-33560	169	34	few	few	ADJ
bracis-33560	169	35	repetitions	repetition	NOUN
bracis-33560	169	36	,	,	PUNCT
bracis-33560	169	37	and	and	CCONJ
bracis-33560	169	38	with	with	ADP
bracis-33560	169	39	semantically	semantically	ADV
bracis-33560	169	40	relevant	relevant	ADJ
bracis-33560	169	41	content	content	NOUN
bracis-33560	169	42	.	.	PUNCT
bracis-33560	170	1	it	it	PRON
bracis-33560	170	2	is	be	AUX
bracis-33560	170	3	worth	worth	ADJ
bracis-33560	170	4	noting	note	VERB
bracis-33560	170	5	that	that	SCONJ
bracis-33560	170	6	we	we	PRON
bracis-33560	170	7	also	also	ADV
bracis-33560	170	8	evaluated	evaluate	VERB
bracis-33560	170	9	the	the	DET
bracis-33560	170	10	number	number	NOUN
bracis-33560	170	11	of	of	ADP
bracis-33560	170	12	potentially	potentially	ADV
bracis-33560	170	13	toxic	toxic	ADJ
bracis-33560	170	14	words	word	NOUN
bracis-33560	170	15	and	and	CCONJ
bracis-33560	170	16	possible	possible	ADJ
bracis-33560	170	17	biases	bias	NOUN
bracis-33560	170	18	contained	contain	VERB
bracis-33560	170	19	in	in	ADP
bracis-33560	170	20	the	the	DET
bracis-33560	170	21	texts	text	NOUN
bracis-33560	170	22	.	.	PUNCT
bracis-33560	171	1	table	table	NOUN
bracis-33560	171	2	 	 	SPACE
bracis-33560	171	3	1	1	NUM
bracis-33560	171	4	shows	show	VERB
bracis-33560	171	5	the	the	DET
bracis-33560	171	6	optimal	optimal	ADJ
bracis-33560	171	7	configuration	configuration	NOUN
bracis-33560	171	8	to	to	PART
bracis-33560	171	9	obtain	obtain	VERB
bracis-33560	171	10	texts	text	NOUN
bracis-33560	171	11	that	that	PRON
bracis-33560	171	12	exceed	exceed	VERB
bracis-33560	171	13	our	our	PRON
bracis-33560	171	14	minimum	minimum	ADJ
bracis-33560	171	15	quality	quality	NOUN
bracis-33560	171	16	requirements	requirement	NOUN
bracis-33560	171	17	.	.	PUNCT
bracis-33560	172	1	table	table	NOUN
bracis-33560	172	2	1	1	NUM
bracis-33560	172	3	.	.	PUNCT
bracis-33560	172	4	double	double	ADJ
bracis-33560	172	5	-	-	PUNCT
bracis-33560	172	6	pipeline	pipeline	NOUN
bracis-33560	172	7	final	final	ADJ
bracis-33560	172	8	configuration.full	configuration.full	NOUN
bracis-33560	172	9	size	size	NOUN
bracis-33560	172	10	table	table	NOUN
bracis-33560	172	11	5	5	NUM
bracis-33560	172	12	results	result	VERB
bracis-33560	172	13	the	the	DET
bracis-33560	172	14	created	create	VERB
bracis-33560	172	15	corpus	corpus	NOUN
bracis-33560	172	16	was	be	AUX
bracis-33560	172	17	evaluated	evaluate	VERB
bracis-33560	172	18	concerning	concern	VERB
bracis-33560	172	19	five	five	NUM
bracis-33560	172	20	groups	group	NOUN
bracis-33560	172	21	of	of	ADP
bracis-33560	172	22	requirements	requirement	NOUN
bracis-33560	172	23	.	.	PUNCT
bracis-33560	173	1	the	the	DET
bracis-33560	173	2	first	first	ADJ
bracis-33560	173	3	requirement	requirement	NOUN
bracis-33560	173	4	is	be	AUX
bracis-33560	173	5	the	the	DET
bracis-33560	173	6	created	create	VERB
bracis-33560	173	7	corpus	corpus	NOUN
bracis-33560	173	8	must	must	AUX
bracis-33560	173	9	be	be	AUX
bracis-33560	173	10	larger	large	ADJ
bracis-33560	173	11	than	than	ADP
bracis-33560	173	12	the	the	DET
bracis-33560	173	13	existing	exist	VERB
bracis-33560	173	14	corpus	corpus	NOUN
bracis-33560	173	15	for	for	ADP
bracis-33560	173	16	the	the	DET
bracis-33560	173	17	portuguese	portuguese	ADJ
bracis-33560	173	18	language	language	NOUN
bracis-33560	173	19	(	(	PUNCT
bracis-33560	173	20	see	see	VERB
bracis-33560	173	21	subsect	subsect	NOUN
bracis-33560	173	22	.	.	PUNCT
bracis-33560	173	23	 	 	SPACE
bracis-33560	173	24	5.1	5.1	NUM
bracis-33560	173	25	)	)	PUNCT
bracis-33560	173	26	.	.	PUNCT
bracis-33560	174	1	the	the	DET
bracis-33560	174	2	second	second	ADJ
bracis-33560	174	3	requirement	requirement	NOUN
bracis-33560	174	4	is	be	AUX
bracis-33560	174	5	the	the	DET
bracis-33560	174	6	corpus	corpus	NOUN
bracis-33560	174	7	must	must	AUX
bracis-33560	174	8	be	be	AUX
bracis-33560	174	9	diverse	diverse	ADJ
bracis-33560	174	10	,	,	PUNCT
bracis-33560	174	11	i.e.	i.e.	X
bracis-33560	174	12	,	,	PUNCT
bracis-33560	174	13	containing	contain	VERB
bracis-33560	174	14	data	datum	NOUN
bracis-33560	174	15	from	from	ADP
bracis-33560	174	16	different	different	ADJ
bracis-33560	174	17	sources	source	NOUN
bracis-33560	174	18	(	(	PUNCT
bracis-33560	174	19	see	see	VERB
bracis-33560	174	20	subsect	subsect	NOUN
bracis-33560	174	21	.	.	PUNCT
bracis-33560	174	22	 	 	SPACE
bracis-33560	174	23	5.2	5.2	NUM
bracis-33560	174	24	)	)	PUNCT
bracis-33560	174	25	.	.	PUNCT
bracis-33560	175	1	the	the	DET
bracis-33560	175	2	third	third	ADJ
bracis-33560	175	3	requirement	requirement	NOUN
bracis-33560	175	4	is	be	AUX
bracis-33560	175	5	the	the	DET
bracis-33560	175	6	corpus	corpus	NOUN
bracis-33560	175	7	covers	cover	VERB
bracis-33560	175	8	the	the	DET
bracis-33560	175	9	most	most	ADV
bracis-33560	175	10	recent	recent	ADJ
bracis-33560	175	11	to	to	ADP
bracis-33560	175	12	the	the	DET
bracis-33560	175	13	least	least	ADJ
bracis-33560	175	14	recent	recent	ADJ
bracis-33560	175	15	information	information	NOUN
bracis-33560	175	16	(	(	PUNCT
bracis-33560	175	17	see	see	VERB
bracis-33560	175	18	subsect	subsect	NOUN
bracis-33560	175	19	.	.	PUNCT
bracis-33560	175	20	 	 	SPACE
bracis-33560	175	21	5.3	5.3	NUM
bracis-33560	175	22	)	)	PUNCT
bracis-33560	175	23	.	.	PUNCT
bracis-33560	176	1	the	the	DET
bracis-33560	176	2	fourth	fourth	ADJ
bracis-33560	176	3	requirement	requirement	NOUN
bracis-33560	176	4	is	be	AUX
bracis-33560	176	5	that	that	SCONJ
bracis-33560	176	6	the	the	DET
bracis-33560	176	7	corpus	corpus	NOUN
bracis-33560	176	8	presents	present	VERB
bracis-33560	176	9	high	high	ADJ
bracis-33560	176	10	-	-	PUNCT
bracis-33560	176	11	quality	quality	NOUN
bracis-33560	176	12	text	text	NOUN
bracis-33560	176	13	indicators	indicator	NOUN
bracis-33560	176	14	(	(	PUNCT
bracis-33560	176	15	see	see	VERB
bracis-33560	176	16	subsect	subsect	NOUN
bracis-33560	176	17	.	.	PUNCT
bracis-33560	176	18	 	 	SPACE
bracis-33560	176	19	5.4	5.4	NUM
bracis-33560	176	20	)	)	PUNCT
bracis-33560	176	21	.	.	PUNCT
bracis-33560	177	1	finally	finally	ADV
bracis-33560	177	2	,	,	PUNCT
bracis-33560	177	3	the	the	DET
bracis-33560	177	4	fifth	fifth	ADJ
bracis-33560	177	5	requirement	requirement	NOUN
bracis-33560	177	6	is	be	AUX
bracis-33560	177	7	that	that	SCONJ
bracis-33560	177	8	the	the	DET
bracis-33560	177	9	corpus	corpus	NOUN
bracis-33560	177	10	avoids	avoid	VERB
bracis-33560	177	11	introducing	introduce	VERB
bracis-33560	177	12	or	or	CCONJ
bracis-33560	177	13	increasing	increase	VERB
bracis-33560	177	14	bias	bias	NOUN
bracis-33560	177	15	.	.	PUNCT
bracis-33560	178	1	due	due	ADP
bracis-33560	178	2	to	to	ADP
bracis-33560	178	3	the	the	DET
bracis-33560	178	4	large	large	ADJ
bracis-33560	178	5	corpus	corpus	NOUN
bracis-33560	178	6	size	size	NOUN
bracis-33560	178	7	,	,	PUNCT
bracis-33560	178	8	subsects	subsect	NOUN
bracis-33560	178	9	.	.	PUNCT
bracis-33560	178	10	 	 	SPACE
bracis-33560	178	11	5.4	5.4	NUM
bracis-33560	178	12	and	and	CCONJ
bracis-33560	178	13	 	 	SPACE
bracis-33560	178	14	5.5	5.5	NUM
bracis-33560	178	15	utilize	utilize	VERB
bracis-33560	178	16	a	a	DET
bracis-33560	178	17	10	10	NUM
bracis-33560	178	18	%	%	NOUN
bracis-33560	178	19	randomly	randomly	ADV
bracis-33560	178	20	generated	generate	VERB
bracis-33560	178	21	sample	sample	NOUN
bracis-33560	178	22	to	to	PART
bracis-33560	178	23	present	present	VERB
bracis-33560	178	24	the	the	DET
bracis-33560	178	25	results	result	NOUN
bracis-33560	178	26	.	.	PUNCT
bracis-33560	179	1	consequently	consequently	ADV
bracis-33560	179	2	,	,	PUNCT
bracis-33560	179	3	fig	fig	NOUN
bracis-33560	179	4	.	.	PUNCT
bracis-33560	179	5	 	 	SPACE
bracis-33560	180	1	4	4	NUM
bracis-33560	180	2	 	 	SPACE
bracis-33560	180	3	and	and	CCONJ
bracis-33560	180	4	 	 	SPACE
bracis-33560	180	5	table	table	NOUN
bracis-33560	180	6	 	 	SPACE
bracis-33560	180	7	3	3	NUM
bracis-33560	180	8	were	be	AUX
bracis-33560	180	9	created	create	VERB
bracis-33560	180	10	based	base	VERB
bracis-33560	180	11	on	on	ADP
bracis-33560	180	12	this	this	DET
bracis-33560	180	13	sample	sample	NOUN
bracis-33560	180	14	size	size	NOUN
bracis-33560	180	15	.	.	PUNCT
bracis-33560	181	1	5.1	5.1	NUM
bracis-33560	181	2	corpus	corpus	NOUN
bracis-33560	181	3	size	size	NOUN
bracis-33560	181	4	the	the	DET
bracis-33560	181	5	first	first	ADJ
bracis-33560	181	6	requirement	requirement	NOUN
bracis-33560	181	7	evaluated	evaluate	VERB
bracis-33560	181	8	was	be	AUX
bracis-33560	181	9	the	the	DET
bracis-33560	181	10	corpus	corpus	NOUN
bracis-33560	181	11	size	size	NOUN
bracis-33560	181	12	.	.	PUNCT
bracis-33560	182	1	we	we	PRON
bracis-33560	182	2	collected	collect	VERB
bracis-33560	182	3	terabytes	terabyte	NOUN
bracis-33560	182	4	of	of	ADP
bracis-33560	182	5	data	datum	NOUN
bracis-33560	182	6	from	from	ADP
bracis-33560	182	7	different	different	ADJ
bracis-33560	182	8	cc	cc	NOUN
bracis-33560	182	9	dumps	dump	NOUN
bracis-33560	182	10	.	.	PUNCT
bracis-33560	183	1	each	each	DET
bracis-33560	183	2	dump	dump	NOUN
bracis-33560	183	3	has	have	VERB
bracis-33560	183	4	approximately	approximately	ADV
bracis-33560	183	5	0.2	0.2	NUM
bracis-33560	183	6	%	%	NOUN
bracis-33560	183	7	of	of	ADP
bracis-33560	183	8	texts	text	NOUN
bracis-33560	183	9	in	in	ADP
bracis-33560	183	10	portuguese	portuguese	PROPN
bracis-33560	183	11	.	.	PUNCT
bracis-33560	184	1	it	it	PRON
bracis-33560	184	2	is	be	AUX
bracis-33560	184	3	worth	worth	ADJ
bracis-33560	184	4	noting	note	VERB
bracis-33560	184	5	that	that	SCONJ
bracis-33560	184	6	the	the	DET
bracis-33560	184	7	documents	document	NOUN
bracis-33560	184	8	may	may	AUX
bracis-33560	184	9	be	be	AUX
bracis-33560	184	10	of	of	ADP
bracis-33560	184	11	poor	poor	ADJ
bracis-33560	184	12	quality	quality	NOUN
bracis-33560	184	13	,	,	PUNCT
bracis-33560	184	14	contain	contain	VERB
bracis-33560	184	15	inappropriate	inappropriate	ADJ
bracis-33560	184	16	or	or	CCONJ
bracis-33560	184	17	biased	biased	ADJ
bracis-33560	184	18	content	content	NOUN
bracis-33560	184	19	,	,	PUNCT
bracis-33560	184	20	and	and	CCONJ
bracis-33560	184	21	be	be	AUX
bracis-33560	184	22	duplicated	duplicate	VERB
bracis-33560	184	23	due	due	ADP
bracis-33560	184	24	to	to	ADP
bracis-33560	184	25	the	the	DET
bracis-33560	184	26	cc	cc	PROPN
bracis-33560	184	27	not	not	PART
bracis-33560	184	28	filtering	filter	VERB
bracis-33560	184	29	the	the	DET
bracis-33560	184	30	data	datum	NOUN
bracis-33560	184	31	.	.	PUNCT
bracis-33560	185	1	therefore	therefore	ADV
bracis-33560	185	2	,	,	PUNCT
bracis-33560	185	3	a	a	DET
bracis-33560	185	4	corpus	corpus	NOUN
bracis-33560	185	5	cleaning	clean	VERB
bracis-33560	185	6	step	step	NOUN
bracis-33560	185	7	was	be	AUX
bracis-33560	185	8	necessary	necessary	ADJ
bracis-33560	185	9	to	to	PART
bracis-33560	185	10	ensure	ensure	VERB
bracis-33560	185	11	that	that	SCONJ
bracis-33560	185	12	the	the	DET
bracis-33560	185	13	final	final	ADJ
bracis-33560	185	14	corpus	corpus	NOUN
bracis-33560	185	15	was	be	AUX
bracis-33560	185	16	composed	compose	VERB
bracis-33560	185	17	only	only	ADV
bracis-33560	185	18	of	of	ADP
bracis-33560	185	19	non	non	ADJ
bracis-33560	185	20	-	-	ADJ
bracis-33560	185	21	duplicated	duplicate	VERB
bracis-33560	185	22	documents	document	NOUN
bracis-33560	185	23	to	to	PART
bracis-33560	185	24	respect	respect	VERB
bracis-33560	185	25	quality	quality	NOUN
bracis-33560	185	26	criteria	criterion	NOUN
bracis-33560	185	27	.	.	PUNCT
bracis-33560	186	1	at	at	ADP
bracis-33560	186	2	the	the	DET
bracis-33560	186	3	end	end	NOUN
bracis-33560	186	4	of	of	ADP
bracis-33560	186	5	this	this	DET
bracis-33560	186	6	process	process	NOUN
bracis-33560	186	7	,	,	PUNCT
bracis-33560	186	8	we	we	PRON
bracis-33560	186	9	obtained	obtain	VERB
bracis-33560	186	10	a	a	DET
bracis-33560	186	11	corpus	corpus	NOUN
bracis-33560	186	12	of	of	ADP
bracis-33560	186	13	100	100	NUM
bracis-33560	186	14	gb	gb	NOUN
bracis-33560	186	15	.	.	PUNCT
bracis-33560	186	16	table	table	NOUN
bracis-33560	186	17	 	 	SPACE
bracis-33560	186	18	2	2	NUM
bracis-33560	186	19	presents	present	VERB
bracis-33560	186	20	the	the	DET
bracis-33560	186	21	created	create	VERB
bracis-33560	186	22	corpus	corpus	NOUN
bracis-33560	186	23	statistics	statistic	NOUN
bracis-33560	186	24	alongside	alongside	ADP
bracis-33560	186	25	other	other	ADJ
bracis-33560	186	26	portuguese	portuguese	PROPN
bracis-33560	186	27	corpora	corpora	PROPN
bracis-33560	186	28	.	.	PUNCT
bracis-33560	187	1	aroeira	aroeira	PROPN
bracis-33560	187	2	surpasses	surpass	VERB
bracis-33560	187	3	brwac	brwac	PROPN
bracis-33560	187	4	 	 	SPACE
bracis-33560	187	5	[	[	X
bracis-33560	187	6	35	35	NUM
bracis-33560	187	7	]	]	PUNCT
bracis-33560	187	8	and	and	CCONJ
bracis-33560	187	9	carolina	carolina	PROPN
bracis-33560	187	10	1.2	1.2	NUM
bracis-33560	187	11	ada	ada	PROPN
bracis-33560	187	12	 	 	SPACE
bracis-33560	188	1	[	[	X
bracis-33560	188	2	11	11	NUM
bracis-33560	188	3	]	]	PUNCT
bracis-33560	188	4	in	in	ADP
bracis-33560	188	5	size	size	NOUN
bracis-33560	188	6	,	,	PUNCT
bracis-33560	188	7	document	document	NOUN
bracis-33560	188	8	quantity	quantity	NOUN
bracis-33560	188	9	,	,	PUNCT
bracis-33560	188	10	and	and	CCONJ
bracis-33560	188	11	token	token	ADJ
bracis-33560	188	12	number	number	NOUN
bracis-33560	188	13	.	.	PUNCT
bracis-33560	189	1	thus	thus	ADV
bracis-33560	189	2	,	,	PUNCT
bracis-33560	189	3	our	our	PRON
bracis-33560	189	4	corpus	corpus	NOUN
bracis-33560	189	5	is	be	AUX
bracis-33560	189	6	potentially	potentially	ADV
bracis-33560	189	7	a	a	DET
bracis-33560	189	8	more	more	ADV
bracis-33560	189	9	diverse	diverse	ADJ
bracis-33560	189	10	resource	resource	NOUN
bracis-33560	189	11	regarding	regard	VERB
bracis-33560	189	12	texts	text	NOUN
bracis-33560	189	13	and	and	CCONJ
bracis-33560	189	14	tokens	token	NOUN
bracis-33560	189	15	than	than	ADP
bracis-33560	189	16	the	the	DET
bracis-33560	189	17	available	available	ADJ
bracis-33560	189	18	resources	resource	NOUN
bracis-33560	189	19	.	.	PUNCT
bracis-33560	190	1	table	table	NOUN
bracis-33560	190	2	2	2	NUM
bracis-33560	190	3	.	.	PUNCT
bracis-33560	190	4	corpora	corpora	NOUN
bracis-33560	190	5	size	size	NOUN
bracis-33560	190	6	comparison.full	comparison.full	ADJ
bracis-33560	190	7	size	size	NOUN
bracis-33560	190	8	table	table	NOUN
bracis-33560	190	9	5.2	5.2	NUM
bracis-33560	190	10	knowledge	knowledge	NOUN
bracis-33560	190	11	domains	domain	NOUN
bracis-33560	190	12	the	the	DET
bracis-33560	190	13	second	second	ADJ
bracis-33560	190	14	requirement	requirement	NOUN
bracis-33560	190	15	evaluated	evaluate	VERB
bracis-33560	190	16	was	be	AUX
bracis-33560	190	17	the	the	DET
bracis-33560	190	18	distribution	distribution	NOUN
bracis-33560	190	19	of	of	ADP
bracis-33560	190	20	knowledge	knowledge	NOUN
bracis-33560	190	21	domains	domain	NOUN
bracis-33560	190	22	within	within	ADP
bracis-33560	190	23	the	the	DET
bracis-33560	190	24	created	create	VERB
bracis-33560	190	25	corpus	corpus	NOUN
bracis-33560	190	26	.	.	PUNCT
bracis-33560	191	1	a	a	DET
bracis-33560	191	2	mapping	mapping	NOUN
bracis-33560	191	3	of	of	ADP
bracis-33560	191	4	different	different	ADJ
bracis-33560	191	5	urls	url	NOUN
bracis-33560	191	6	to	to	ADP
bracis-33560	191	7	their	their	PRON
bracis-33560	191	8	respective	respective	ADJ
bracis-33560	191	9	knowledge	knowledge	NOUN
bracis-33560	191	10	domains	domain	NOUN
bracis-33560	191	11	was	be	AUX
bracis-33560	191	12	conducted	conduct	VERB
bracis-33560	191	13	.	.	PUNCT
bracis-33560	192	1	each	each	DET
bracis-33560	192	2	base	base	NOUN
bracis-33560	192	3	url	url	NOUN
bracis-33560	192	4	was	be	AUX
bracis-33560	192	5	verified	verify	VERB
bracis-33560	192	6	against	against	ADP
bracis-33560	192	7	a	a	DET
bracis-33560	192	8	dictionary	dictionary	NOUN
bracis-33560	192	9	.	.	PUNCT
bracis-33560	193	1	when	when	SCONJ
bracis-33560	193	2	there	there	PRON
bracis-33560	193	3	were	be	VERB
bracis-33560	193	4	no	no	DET
bracis-33560	193	5	matches	match	NOUN
bracis-33560	193	6	,	,	PUNCT
bracis-33560	193	7	keywords	keyword	NOUN
bracis-33560	193	8	were	be	AUX
bracis-33560	193	9	used	use	VERB
bracis-33560	193	10	to	to	PART
bracis-33560	193	11	determine	determine	VERB
bracis-33560	193	12	the	the	DET
bracis-33560	193	13	document	document	NOUN
bracis-33560	193	14	’s	’s	PART
bracis-33560	193	15	domain	domain	NOUN
bracis-33560	193	16	 	 	SPACE
bracis-33560	193	17	(	(	PUNCT
bracis-33560	193	18	subsect	subsect	NOUN
bracis-33560	193	19	.	.	PUNCT
bracis-33560	193	20	 	 	SPACE
bracis-33560	193	21	3.9	3.9	NUM
bracis-33560	193	22	)	)	PUNCT
bracis-33560	193	23	.	.	PUNCT
bracis-33560	194	1	figure	figure	VERB
bracis-33560	194	2	 	 	SPACE
bracis-33560	194	3	2	2	NUM
bracis-33560	194	4	illustrates	illustrate	VERB
bracis-33560	194	5	the	the	DET
bracis-33560	194	6	document	document	NOUN
bracis-33560	194	7	distribution	distribution	NOUN
bracis-33560	194	8	across	across	ADP
bracis-33560	194	9	these	these	DET
bracis-33560	194	10	domains	domain	NOUN
bracis-33560	194	11	.	.	PUNCT
bracis-33560	195	1	the	the	DET
bracis-33560	195	2	complexity	complexity	NOUN
bracis-33560	195	3	of	of	ADP
bracis-33560	195	4	the	the	DET
bracis-33560	195	5	corpus	corpus	NOUN
bracis-33560	195	6	strongly	strongly	ADV
bracis-33560	195	7	correlates	correlate	VERB
bracis-33560	195	8	with	with	ADP
bracis-33560	195	9	downstream	downstream	ADJ
bracis-33560	195	10	data	datum	NOUN
bracis-33560	195	11	performance	performance	NOUN
bracis-33560	195	12	 	 	SPACE
bracis-33560	195	13	[	[	X
bracis-33560	195	14	3	3	NUM
bracis-33560	195	15	]	]	PUNCT
bracis-33560	195	16	.	.	PUNCT
bracis-33560	196	1	therefore	therefore	ADV
bracis-33560	196	2	,	,	PUNCT
bracis-33560	196	3	an	an	DET
bracis-33560	196	4	extensive	extensive	ADJ
bracis-33560	196	5	representation	representation	NOUN
bracis-33560	196	6	of	of	ADP
bracis-33560	196	7	knowledge	knowledge	NOUN
bracis-33560	196	8	domains	domain	NOUN
bracis-33560	196	9	can	can	AUX
bracis-33560	196	10	contribute	contribute	VERB
bracis-33560	196	11	to	to	ADP
bracis-33560	196	12	the	the	DET
bracis-33560	196	13	generation	generation	NOUN
bracis-33560	196	14	of	of	ADP
bracis-33560	196	15	more	more	ADV
bracis-33560	196	16	robust	robust	ADJ
bracis-33560	196	17	models	model	NOUN
bracis-33560	196	18	,	,	PUNCT
bracis-33560	196	19	potentially	potentially	ADV
bracis-33560	196	20	improving	improve	VERB
bracis-33560	196	21	in	in	ADP
bracis-33560	196	22	-	-	PUNCT
bracis-33560	196	23	context	context	NOUN
bracis-33560	196	24	few	few	ADJ
bracis-33560	196	25	-	-	PUNCT
bracis-33560	196	26	shot	shot	NOUN
bracis-33560	196	27	learning	learning	NOUN
bracis-33560	196	28	performance	performance	NOUN
bracis-33560	196	29	 	 	SPACE
bracis-33560	197	1	[	[	X
bracis-33560	197	2	32	32	NUM
bracis-33560	197	3	]	]	PUNCT
bracis-33560	197	4	.	.	PUNCT
bracis-33560	198	1	most	most	ADJ
bracis-33560	198	2	documents	document	NOUN
bracis-33560	198	3	could	could	AUX
bracis-33560	198	4	not	not	PART
bracis-33560	198	5	be	be	AUX
bracis-33560	198	6	assigned	assign	VERB
bracis-33560	198	7	to	to	ADP
bracis-33560	198	8	a	a	DET
bracis-33560	198	9	specific	specific	ADJ
bracis-33560	198	10	domain	domain	NOUN
bracis-33560	198	11	and	and	CCONJ
bracis-33560	198	12	are	be	AUX
bracis-33560	198	13	marked	mark	VERB
bracis-33560	198	14	as	as	ADP
bracis-33560	198	15	nr	nr	PROPN
bracis-33560	198	16	(	(	PUNCT
bracis-33560	198	17	not	not	PART
bracis-33560	198	18	recognized	recognize	VERB
bracis-33560	198	19	)	)	PUNCT
bracis-33560	198	20	.	.	PUNCT
bracis-33560	199	1	among	among	ADP
bracis-33560	199	2	those	those	PRON
bracis-33560	199	3	that	that	PRON
bracis-33560	199	4	were	be	AUX
bracis-33560	199	5	categorized	categorize	VERB
bracis-33560	199	6	,	,	PUNCT
bracis-33560	199	7	blog	blog	NOUN
bracis-33560	199	8	posts	post	NOUN
bracis-33560	199	9	and	and	CCONJ
bracis-33560	199	10	news	news	NOUN
bracis-33560	199	11	articles	article	NOUN
bracis-33560	199	12	were	be	AUX
bracis-33560	199	13	the	the	DET
bracis-33560	199	14	most	most	ADV
bracis-33560	199	15	frequent	frequent	ADJ
bracis-33560	199	16	,	,	PUNCT
bracis-33560	199	17	although	although	SCONJ
bracis-33560	199	18	other	other	ADJ
bracis-33560	199	19	categories	category	NOUN
bracis-33560	199	20	such	such	ADJ
bracis-33560	199	21	as	as	ADP
bracis-33560	199	22	institutional	institutional	ADJ
bracis-33560	199	23	texts	text	NOUN
bracis-33560	199	24	,	,	PUNCT
bracis-33560	199	25	e	e	NOUN
bracis-33560	199	26	-	-	NOUN
bracis-33560	199	27	commerce	commerce	NOUN
bracis-33560	199	28	,	,	PUNCT
bracis-33560	199	29	and	and	CCONJ
bracis-33560	199	30	internet	internet	NOUN
bracis-33560	199	31	forums	forum	NOUN
bracis-33560	199	32	were	be	AUX
bracis-33560	199	33	also	also	ADV
bracis-33560	199	34	found	find	VERB
bracis-33560	199	35	.	.	PUNCT
bracis-33560	200	1	knowledge	knowledge	NOUN
bracis-33560	200	2	domains	domain	NOUN
bracis-33560	200	3	are	be	AUX
bracis-33560	200	4	essential	essential	ADJ
bracis-33560	200	5	for	for	ADP
bracis-33560	200	6	evaluating	evaluate	VERB
bracis-33560	200	7	the	the	DET
bracis-33560	200	8	quality	quality	NOUN
bracis-33560	200	9	of	of	ADP
bracis-33560	200	10	the	the	DET
bracis-33560	200	11	data	datum	NOUN
bracis-33560	200	12	in	in	ADP
bracis-33560	200	13	the	the	DET
bracis-33560	200	14	corpus	corpus	NOUN
bracis-33560	200	15	and	and	CCONJ
bracis-33560	200	16	for	for	ADP
bracis-33560	200	17	filtering	filter	VERB
bracis-33560	200	18	data	datum	NOUN
bracis-33560	200	19	used	use	VERB
bracis-33560	200	20	in	in	ADP
bracis-33560	200	21	the	the	DET
bracis-33560	200	22	pre	pre	ADJ
bracis-33560	200	23	-	-	ADJ
bracis-33560	200	24	training	training	ADJ
bracis-33560	200	25	phase	phase	NOUN
bracis-33560	200	26	of	of	ADP
bracis-33560	200	27	domain	domain	NOUN
bracis-33560	200	28	-	-	PUNCT
bracis-33560	200	29	specific	specific	ADJ
bracis-33560	200	30	language	language	NOUN
bracis-33560	200	31	models	model	NOUN
bracis-33560	200	32	.	.	PUNCT
bracis-33560	201	1	5.3	5.3	NUM
bracis-33560	201	2	distribution	distribution	NOUN
bracis-33560	201	3	of	of	ADP
bracis-33560	201	4	 	 	SPACE
bracis-33560	201	5	documents	document	NOUN
bracis-33560	201	6	over	over	ADP
bracis-33560	201	7	 	 	SPACE
bracis-33560	201	8	time	time	NOUN
bracis-33560	201	9	our	our	PRON
bracis-33560	201	10	third	third	ADJ
bracis-33560	201	11	analysis	analysis	NOUN
bracis-33560	201	12	is	be	AUX
bracis-33560	201	13	the	the	DET
bracis-33560	201	14	distribution	distribution	NOUN
bracis-33560	201	15	of	of	ADP
bracis-33560	201	16	the	the	DET
bracis-33560	201	17	corpus	corpus	NOUN
bracis-33560	201	18	documents	document	NOUN
bracis-33560	201	19	over	over	ADP
bracis-33560	201	20	time	time	NOUN
bracis-33560	201	21	.	.	PUNCT
bracis-33560	202	1	this	this	DET
bracis-33560	202	2	temporal	temporal	ADJ
bracis-33560	202	3	analysis	analysis	NOUN
bracis-33560	202	4	is	be	AUX
bracis-33560	202	5	important	important	ADJ
bracis-33560	202	6	to	to	PART
bracis-33560	202	7	identify	identify	VERB
bracis-33560	202	8	possible	possible	ADJ
bracis-33560	202	9	temporal	temporal	ADJ
bracis-33560	202	10	biases	bias	NOUN
bracis-33560	202	11	such	such	ADJ
bracis-33560	202	12	as	as	ADP
bracis-33560	202	13	outdated	outdated	ADJ
bracis-33560	202	14	texts	text	NOUN
bracis-33560	202	15	.	.	PUNCT
bracis-33560	203	1	our	our	PRON
bracis-33560	203	2	corpus	corpus	NOUN
bracis-33560	203	3	presents	present	VERB
bracis-33560	203	4	a	a	DET
bracis-33560	203	5	recent	recent	ADJ
bracis-33560	203	6	data	datum	NOUN
bracis-33560	203	7	distribution	distribution	NOUN
bracis-33560	203	8	,	,	PUNCT
bracis-33560	203	9	which	which	PRON
bracis-33560	203	10	indicates	indicate	VERB
bracis-33560	203	11	more	more	ADJ
bracis-33560	203	12	up	up	ADP
bracis-33560	203	13	-	-	PUNCT
bracis-33560	203	14	to	to	ADP
bracis-33560	203	15	-	-	PUNCT
bracis-33560	203	16	date	date	NOUN
bracis-33560	203	17	texts	text	NOUN
bracis-33560	203	18	.	.	PUNCT
bracis-33560	204	1	figure	figure	NOUN
bracis-33560	204	2	 	 	SPACE
bracis-33560	204	3	3	3	NUM
bracis-33560	204	4	illustrates	illustrate	VERB
bracis-33560	204	5	that	that	SCONJ
bracis-33560	204	6	our	our	PRON
bracis-33560	204	7	data	datum	NOUN
bracis-33560	204	8	set	set	VERB
bracis-33560	204	9	comprises	comprise	VERB
bracis-33560	204	10	documents	document	NOUN
bracis-33560	204	11	spanning	span	VERB
bracis-33560	204	12	up	up	ADP
bracis-33560	204	13	to	to	PART
bracis-33560	204	14	7	7	NUM
bracis-33560	204	15	years	year	NOUN
bracis-33560	204	16	,	,	PUNCT
bracis-33560	204	17	beginning	begin	VERB
bracis-33560	204	18	in	in	ADP
bracis-33560	204	19	2017	2017	NUM
bracis-33560	204	20	.	.	PUNCT
bracis-33560	205	1	the	the	DET
bracis-33560	205	2	bulk	bulk	NOUN
bracis-33560	205	3	of	of	ADP
bracis-33560	205	4	the	the	DET
bracis-33560	205	5	data	data	NOUN
bracis-33560	205	6	is	be	AUX
bracis-33560	205	7	from	from	ADP
bracis-33560	205	8	2017	2017	NUM
bracis-33560	205	9	to	to	ADP
bracis-33560	205	10	2019	2019	NUM
bracis-33560	205	11	,	,	PUNCT
bracis-33560	205	12	but	but	CCONJ
bracis-33560	205	13	a	a	DET
bracis-33560	205	14	notable	notable	ADJ
bracis-33560	205	15	portion	portion	NOUN
bracis-33560	205	16	of	of	ADP
bracis-33560	205	17	recent	recent	ADJ
bracis-33560	205	18	data	datum	NOUN
bracis-33560	205	19	is	be	AUX
bracis-33560	205	20	from	from	ADP
bracis-33560	205	21	2021	2021	NUM
bracis-33560	205	22	,	,	PUNCT
bracis-33560	205	23	2022	2022	NUM
bracis-33560	205	24	,	,	PUNCT
bracis-33560	205	25	and	and	CCONJ
bracis-33560	205	26	2023	2023	NUM
bracis-33560	205	27	.	.	PUNCT
bracis-33560	206	1	this	this	DET
bracis-33560	206	2	distribution	distribution	NOUN
bracis-33560	206	3	meets	meet	VERB
bracis-33560	206	4	the	the	DET
bracis-33560	206	5	need	need	NOUN
bracis-33560	206	6	for	for	ADP
bracis-33560	206	7	both	both	CCONJ
bracis-33560	206	8	recent	recent	ADJ
bracis-33560	206	9	and	and	CCONJ
bracis-33560	206	10	extensive	extensive	ADJ
bracis-33560	206	11	data	datum	NOUN
bracis-33560	206	12	.	.	PUNCT
bracis-33560	207	1	as	as	ADP
bracis-33560	207	2	a	a	DET
bracis-33560	207	3	result	result	NOUN
bracis-33560	207	4	,	,	PUNCT
bracis-33560	207	5	models	model	NOUN
bracis-33560	207	6	can	can	AUX
bracis-33560	207	7	train	train	VERB
bracis-33560	207	8	on	on	ADP
bracis-33560	207	9	up	up	ADP
bracis-33560	207	10	-	-	PUNCT
bracis-33560	207	11	to	to	ADP
bracis-33560	207	12	-	-	PUNCT
bracis-33560	207	13	date	date	NOUN
bracis-33560	207	14	information	information	NOUN
bracis-33560	207	15	and	and	CCONJ
bracis-33560	207	16	include	include	VERB
bracis-33560	207	17	recent	recent	ADJ
bracis-33560	207	18	and	and	CCONJ
bracis-33560	207	19	common	common	ADJ
bracis-33560	207	20	terms	term	NOUN
bracis-33560	207	21	used	use	VERB
bracis-33560	207	22	in	in	ADP
bracis-33560	207	23	portuguese	portuguese	PROPN
bracis-33560	207	24	.	.	PUNCT
bracis-33560	208	1	fig	fig	NOUN
bracis-33560	208	2	.	.	PUNCT
bracis-33560	209	1	2	2	X
bracis-33560	209	2	.	.	X
bracis-33560	209	3	distribution	distribution	NOUN
bracis-33560	209	4	of	of	ADP
bracis-33560	209	5	knowledge	knowledge	NOUN
bracis-33560	209	6	domains	domain	NOUN
bracis-33560	209	7	.	.	PUNCT
bracis-33560	210	1	full	full	ADJ
bracis-33560	210	2	size	size	NOUN
bracis-33560	210	3	image	image	NOUN
bracis-33560	210	4	fig	fig	NOUN
bracis-33560	210	5	.	.	PUNCT
bracis-33560	211	1	3	3	X
bracis-33560	211	2	.	.	X
bracis-33560	211	3	distribution	distribution	NOUN
bracis-33560	211	4	of	of	ADP
bracis-33560	211	5	documents	document	NOUN
bracis-33560	211	6	over	over	ADP
bracis-33560	211	7	time	time	NOUN
bracis-33560	211	8	.	.	PUNCT
bracis-33560	212	1	full	full	ADJ
bracis-33560	212	2	size	size	NOUN
bracis-33560	212	3	image	image	NOUN
bracis-33560	212	4	fig	fig	NOUN
bracis-33560	212	5	.	.	PUNCT
bracis-33560	213	1	4	4	X
bracis-33560	213	2	.	.	X
bracis-33560	213	3	quality	quality	NOUN
bracis-33560	213	4	indicators	indicator	NOUN
bracis-33560	213	5	.	.	PUNCT
bracis-33560	214	1	higher	high	ADJ
bracis-33560	214	2	ttr	ttr	NOUN
bracis-33560	214	3	and	and	CCONJ
bracis-33560	214	4	percentage	percentage	NOUN
bracis-33560	214	5	of	of	ADP
bracis-33560	214	6	valid	valid	ADJ
bracis-33560	214	7	words	word	NOUN
bracis-33560	214	8	indicate	indicate	VERB
bracis-33560	214	9	better	well	ADJ
bracis-33560	214	10	corpus	corpus	NOUN
bracis-33560	214	11	quality	quality	NOUN
bracis-33560	214	12	,	,	PUNCT
bracis-33560	214	13	while	while	SCONJ
bracis-33560	214	14	lower	low	ADJ
bracis-33560	214	15	values	value	NOUN
bracis-33560	214	16	for	for	ADP
bracis-33560	214	17	other	other	ADJ
bracis-33560	214	18	indicators	indicator	NOUN
bracis-33560	214	19	also	also	ADV
bracis-33560	214	20	signify	signify	VERB
bracis-33560	214	21	better	well	ADJ
bracis-33560	214	22	quality	quality	NOUN
bracis-33560	214	23	.	.	PUNCT
bracis-33560	215	1	full	full	ADJ
bracis-33560	215	2	size	size	NOUN
bracis-33560	215	3	image	image	NOUN
bracis-33560	215	4	5.4	5.4	NUM
bracis-33560	215	5	quality	quality	NOUN
bracis-33560	215	6	indicators	indicator	NOUN
bracis-33560	215	7	we	we	PRON
bracis-33560	215	8	used	use	VERB
bracis-33560	215	9	ttr	ttr	PROPN
bracis-33560	215	10	value	value	NOUN
bracis-33560	215	11	,	,	PUNCT
bracis-33560	215	12	symbol	symbol	NOUN
bracis-33560	215	13	word	word	NOUN
bracis-33560	215	14	percentage	percentage	NOUN
bracis-33560	215	15	,	,	PUNCT
bracis-33560	215	16	stopword	stopword	NOUN
bracis-33560	215	17	percentage	percentage	NOUN
bracis-33560	215	18	,	,	PUNCT
bracis-33560	215	19	valid	valid	ADJ
bracis-33560	215	20	words	word	NOUN
bracis-33560	215	21	,	,	PUNCT
bracis-33560	215	22	and	and	CCONJ
bracis-33560	215	23	toxic	toxic	ADJ
bracis-33560	215	24	content	content	NOUN
bracis-33560	215	25	as	as	ADP
bracis-33560	215	26	quality	quality	NOUN
bracis-33560	215	27	indicators	indicator	NOUN
bracis-33560	215	28	.	.	PUNCT
bracis-33560	216	1	figure	figure	VERB
bracis-33560	216	2	 	 	SPACE
bracis-33560	216	3	4	4	NUM
bracis-33560	216	4	shows	show	VERB
bracis-33560	216	5	the	the	DET
bracis-33560	216	6	results	result	NOUN
bracis-33560	216	7	obtained	obtain	VERB
bracis-33560	216	8	.	.	PUNCT
bracis-33560	217	1	we	we	PRON
bracis-33560	217	2	have	have	VERB
bracis-33560	217	3	two	two	NUM
bracis-33560	217	4	metrics	metric	NOUN
bracis-33560	217	5	that	that	PRON
bracis-33560	217	6	indicate	indicate	VERB
bracis-33560	217	7	the	the	DET
bracis-33560	217	8	diversity	diversity	NOUN
bracis-33560	217	9	of	of	ADP
bracis-33560	217	10	the	the	DET
bracis-33560	217	11	content	content	NOUN
bracis-33560	217	12	present	present	ADJ
bracis-33560	217	13	in	in	ADP
bracis-33560	217	14	our	our	PRON
bracis-33560	217	15	corpus	corpus	NOUN
bracis-33560	217	16	.	.	PUNCT
bracis-33560	218	1	the	the	DET
bracis-33560	218	2	ttr	ttr	PROPN
bracis-33560	218	3	indicates	indicate	VERB
bracis-33560	218	4	the	the	DET
bracis-33560	218	5	variability	variability	NOUN
bracis-33560	218	6	of	of	ADP
bracis-33560	218	7	tokens	token	NOUN
bracis-33560	218	8	in	in	ADP
bracis-33560	218	9	a	a	DET
bracis-33560	218	10	sentence	sentence	NOUN
bracis-33560	218	11	,	,	PUNCT
bracis-33560	218	12	and	and	CCONJ
bracis-33560	218	13	we	we	PRON
bracis-33560	218	14	aim	aim	VERB
bracis-33560	218	15	for	for	SCONJ
bracis-33560	218	16	this	this	DET
bracis-33560	218	17	value	value	NOUN
bracis-33560	218	18	to	to	PART
bracis-33560	218	19	be	be	AUX
bracis-33560	218	20	as	as	ADV
bracis-33560	218	21	high	high	ADJ
bracis-33560	218	22	as	as	ADP
bracis-33560	218	23	possible	possible	ADJ
bracis-33560	218	24	,	,	PUNCT
bracis-33560	218	25	as	as	SCONJ
bracis-33560	218	26	it	it	PRON
bracis-33560	218	27	is	be	AUX
bracis-33560	218	28	a	a	DET
bracis-33560	218	29	sign	sign	NOUN
bracis-33560	218	30	of	of	ADP
bracis-33560	218	31	texts	text	NOUN
bracis-33560	218	32	composed	compose	VERB
bracis-33560	218	33	of	of	ADP
bracis-33560	218	34	varied	varied	ADJ
bracis-33560	218	35	tokens	token	NOUN
bracis-33560	218	36	with	with	ADP
bracis-33560	218	37	low	low	ADJ
bracis-33560	218	38	word	word	NOUN
bracis-33560	218	39	repetition	repetition	NOUN
bracis-33560	218	40	.	.	PUNCT
bracis-33560	219	1	a	a	DET
bracis-33560	219	2	ttr	ttr	PROPN
bracis-33560	219	3	threshold	threshold	NOUN
bracis-33560	219	4	of	of	ADP
bracis-33560	219	5	0.5	0.5	NUM
bracis-33560	219	6	is	be	AUX
bracis-33560	219	7	a	a	DET
bracis-33560	219	8	quality	quality	NOUN
bracis-33560	219	9	parameter	parameter	NOUN
bracis-33560	219	10	for	for	ADP
bracis-33560	219	11	the	the	DET
bracis-33560	219	12	text	text	NOUN
bracis-33560	219	13	.	.	PUNCT
bracis-33560	220	1	we	we	PRON
bracis-33560	220	2	obtained	obtain	VERB
bracis-33560	220	3	a	a	DET
bracis-33560	220	4	distribution	distribution	NOUN
bracis-33560	220	5	curve	curve	NOUN
bracis-33560	220	6	with	with	ADP
bracis-33560	220	7	the	the	DET
bracis-33560	220	8	first	first	ADJ
bracis-33560	220	9	quartile	quartile	NOUN
bracis-33560	220	10	close	close	ADV
bracis-33560	220	11	to	to	ADP
bracis-33560	220	12	0.5	0.5	NUM
bracis-33560	220	13	ttr	ttr	NOUN
bracis-33560	220	14	,	,	PUNCT
bracis-33560	220	15	a	a	DET
bracis-33560	220	16	median	median	NOUN
bracis-33560	220	17	of	of	ADP
bracis-33560	220	18	0.57	0.57	NUM
bracis-33560	220	19	,	,	PUNCT
bracis-33560	220	20	and	and	CCONJ
bracis-33560	220	21	the	the	DET
bracis-33560	220	22	third	third	ADJ
bracis-33560	220	23	quartile	quartile	NOUN
bracis-33560	220	24	above	above	ADP
bracis-33560	220	25	0.65	0.65	NUM
bracis-33560	220	26	.	.	PUNCT
bracis-33560	221	1	thus	thus	ADV
bracis-33560	221	2	,	,	PUNCT
bracis-33560	221	3	most	most	ADJ
bracis-33560	221	4	of	of	ADP
bracis-33560	221	5	the	the	DET
bracis-33560	221	6	data	datum	NOUN
bracis-33560	221	7	in	in	ADP
bracis-33560	221	8	our	our	PRON
bracis-33560	221	9	corpus	corpus	NOUN
bracis-33560	221	10	reaches	reach	VERB
bracis-33560	221	11	a	a	DET
bracis-33560	221	12	satisfactory	satisfactory	ADJ
bracis-33560	221	13	ttr	ttr	PROPN
bracis-33560	221	14	value	value	NOUN
bracis-33560	221	15	,	,	PUNCT
bracis-33560	221	16	which	which	PRON
bracis-33560	221	17	is	be	AUX
bracis-33560	221	18	a	a	DET
bracis-33560	221	19	strong	strong	ADJ
bracis-33560	221	20	indicator	indicator	NOUN
bracis-33560	221	21	of	of	ADP
bracis-33560	221	22	non	non	ADJ
bracis-33560	221	23	-	-	ADJ
bracis-33560	221	24	repetitive	repetitive	ADJ
bracis-33560	221	25	texts	text	NOUN
bracis-33560	221	26	.	.	PUNCT
bracis-33560	222	1	another	another	DET
bracis-33560	222	2	indicator	indicator	NOUN
bracis-33560	222	3	of	of	ADP
bracis-33560	222	4	variability	variability	NOUN
bracis-33560	222	5	is	be	AUX
bracis-33560	222	6	the	the	DET
bracis-33560	222	7	percentage	percentage	NOUN
bracis-33560	222	8	of	of	ADP
bracis-33560	222	9	valid	valid	ADJ
bracis-33560	222	10	tokens	token	NOUN
bracis-33560	222	11	or	or	CCONJ
bracis-33560	222	12	keywords	keyword	NOUN
bracis-33560	222	13	.	.	PUNCT
bracis-33560	223	1	this	this	DET
bracis-33560	223	2	indicator	indicator	NOUN
bracis-33560	223	3	measures	measure	VERB
bracis-33560	223	4	the	the	DET
bracis-33560	223	5	frequency	frequency	NOUN
bracis-33560	223	6	of	of	ADP
bracis-33560	223	7	contextually	contextually	ADV
bracis-33560	223	8	relevant	relevant	ADJ
bracis-33560	223	9	words	word	NOUN
bracis-33560	223	10	within	within	ADP
bracis-33560	223	11	the	the	DET
bracis-33560	223	12	text	text	NOUN
bracis-33560	223	13	,	,	PUNCT
bracis-33560	223	14	and	and	CCONJ
bracis-33560	223	15	we	we	PRON
bracis-33560	223	16	also	also	ADV
bracis-33560	223	17	want	want	VERB
bracis-33560	223	18	higher	high	ADJ
bracis-33560	223	19	values	value	NOUN
bracis-33560	223	20	for	for	ADP
bracis-33560	223	21	more	more	ADV
bracis-33560	223	22	diverse	diverse	ADJ
bracis-33560	223	23	and	and	CCONJ
bracis-33560	223	24	fluid	fluid	ADJ
bracis-33560	223	25	texts	text	NOUN
bracis-33560	223	26	.	.	PUNCT
bracis-33560	224	1	we	we	PRON
bracis-33560	224	2	achieved	achieve	VERB
bracis-33560	224	3	an	an	DET
bracis-33560	224	4	excellent	excellent	ADJ
bracis-33560	224	5	distribution	distribution	NOUN
bracis-33560	224	6	in	in	ADP
bracis-33560	224	7	this	this	DET
bracis-33560	224	8	indicator	indicator	NOUN
bracis-33560	224	9	,	,	PUNCT
bracis-33560	224	10	with	with	ADP
bracis-33560	224	11	most	most	ADJ
bracis-33560	224	12	data	datum	NOUN
bracis-33560	224	13	above	above	ADP
bracis-33560	224	14	0.7	0.7	NUM
bracis-33560	224	15	.	.	PUNCT
bracis-33560	225	1	this	this	DET
bracis-33560	225	2	distribution	distribution	NOUN
bracis-33560	225	3	value	value	NOUN
bracis-33560	225	4	is	be	AUX
bracis-33560	225	5	another	another	DET
bracis-33560	225	6	strong	strong	ADJ
bracis-33560	225	7	sign	sign	NOUN
bracis-33560	225	8	of	of	ADP
bracis-33560	225	9	lexically	lexically	ADV
bracis-33560	225	10	diverse	diverse	ADJ
bracis-33560	225	11	texts	text	NOUN
bracis-33560	225	12	.	.	PUNCT
bracis-33560	226	1	in	in	ADP
bracis-33560	226	2	contrast	contrast	NOUN
bracis-33560	226	3	,	,	PUNCT
bracis-33560	226	4	the	the	DET
bracis-33560	226	5	metrics	metric	NOUN
bracis-33560	226	6	for	for	ADP
bracis-33560	226	7	the	the	DET
bracis-33560	226	8	percentage	percentage	NOUN
bracis-33560	226	9	of	of	ADP
bracis-33560	226	10	symbols	symbol	NOUN
bracis-33560	226	11	and	and	CCONJ
bracis-33560	226	12	the	the	DET
bracis-33560	226	13	percentage	percentage	NOUN
bracis-33560	226	14	of	of	ADP
bracis-33560	226	15	stopwords	stopword	NOUN
bracis-33560	226	16	are	be	AUX
bracis-33560	226	17	indicators	indicator	NOUN
bracis-33560	226	18	of	of	ADP
bracis-33560	226	19	less	less	ADJ
bracis-33560	226	20	fluid	fluid	ADJ
bracis-33560	226	21	texts	text	NOUN
bracis-33560	226	22	,	,	PUNCT
bracis-33560	226	23	with	with	ADP
bracis-33560	226	24	many	many	ADJ
bracis-33560	226	25	symbols	symbol	NOUN
bracis-33560	226	26	interrupting	interrupt	VERB
bracis-33560	226	27	the	the	DET
bracis-33560	226	28	text	text	NOUN
bracis-33560	226	29	or	or	CCONJ
bracis-33560	226	30	commonly	commonly	ADV
bracis-33560	226	31	used	use	VERB
bracis-33560	226	32	words	word	NOUN
bracis-33560	226	33	that	that	PRON
bracis-33560	226	34	do	do	AUX
bracis-33560	226	35	not	not	PART
bracis-33560	226	36	add	add	VERB
bracis-33560	226	37	semantic	semantic	ADJ
bracis-33560	226	38	value	value	NOUN
bracis-33560	226	39	to	to	ADP
bracis-33560	226	40	the	the	DET
bracis-33560	226	41	documents	document	NOUN
bracis-33560	226	42	(	(	PUNCT
bracis-33560	226	43	e.g.	e.g.	ADV
bracis-33560	226	44	,	,	PUNCT
bracis-33560	226	45	“	"	PUNCT
bracis-33560	226	46	to	to	AUX
bracis-33560	226	47	”	"	PUNCT
bracis-33560	226	48	,	,	PUNCT
bracis-33560	226	49	“	"	PUNCT
bracis-33560	226	50	for	for	ADP
bracis-33560	226	51	”	"	PUNCT
bracis-33560	226	52	,	,	PUNCT
bracis-33560	226	53	or	or	CCONJ
bracis-33560	226	54	“	"	PUNCT
bracis-33560	226	55	the	the	PRON
bracis-33560	226	56	”	"	PUNCT
bracis-33560	226	57	)	)	PUNCT
bracis-33560	226	58	.	.	PUNCT
bracis-33560	227	1	our	our	PRON
bracis-33560	227	2	goal	goal	NOUN
bracis-33560	227	3	is	be	AUX
bracis-33560	227	4	to	to	PART
bracis-33560	227	5	minimize	minimize	VERB
bracis-33560	227	6	these	these	DET
bracis-33560	227	7	metrics	metric	NOUN
bracis-33560	227	8	as	as	ADV
bracis-33560	227	9	much	much	ADV
bracis-33560	227	10	as	as	ADP
bracis-33560	227	11	possible	possible	ADJ
bracis-33560	227	12	.	.	PUNCT
bracis-33560	228	1	we	we	PRON
bracis-33560	228	2	achieved	achieve	VERB
bracis-33560	228	3	our	our	PRON
bracis-33560	228	4	goal	goal	NOUN
bracis-33560	228	5	of	of	ADP
bracis-33560	228	6	reducing	reduce	VERB
bracis-33560	228	7	these	these	DET
bracis-33560	228	8	values	value	NOUN
bracis-33560	228	9	,	,	PUNCT
bracis-33560	228	10	obtaining	obtain	VERB
bracis-33560	228	11	distributions	distribution	NOUN
bracis-33560	228	12	with	with	ADP
bracis-33560	228	13	low	low	ADJ
bracis-33560	228	14	values	value	NOUN
bracis-33560	228	15	,	,	PUNCT
bracis-33560	228	16	with	with	ADP
bracis-33560	228	17	the	the	DET
bracis-33560	228	18	third	third	ADJ
bracis-33560	228	19	quartile	quartile	NOUN
bracis-33560	228	20	below	below	ADP
bracis-33560	228	21	0.2	0.2	NUM
bracis-33560	228	22	in	in	ADP
bracis-33560	228	23	both	both	DET
bracis-33560	228	24	metrics	metric	NOUN
bracis-33560	228	25	.	.	PUNCT
bracis-33560	229	1	this	this	DET
bracis-33560	229	2	result	result	NOUN
bracis-33560	229	3	is	be	AUX
bracis-33560	229	4	a	a	DET
bracis-33560	229	5	strong	strong	ADJ
bracis-33560	229	6	indication	indication	NOUN
bracis-33560	229	7	that	that	SCONJ
bracis-33560	229	8	the	the	DET
bracis-33560	229	9	texts	text	NOUN
bracis-33560	229	10	in	in	ADP
bracis-33560	229	11	the	the	DET
bracis-33560	229	12	corpus	corpus	NOUN
bracis-33560	229	13	are	be	AUX
bracis-33560	229	14	fluid	fluid	ADJ
bracis-33560	229	15	.	.	PUNCT
bracis-33560	230	1	finally	finally	ADV
bracis-33560	230	2	,	,	PUNCT
bracis-33560	230	3	we	we	PRON
bracis-33560	230	4	aim	aim	VERB
bracis-33560	230	5	to	to	PART
bracis-33560	230	6	minimize	minimize	VERB
bracis-33560	230	7	the	the	DET
bracis-33560	230	8	percentage	percentage	NOUN
bracis-33560	230	9	of	of	ADP
bracis-33560	230	10	potentially	potentially	ADV
bracis-33560	230	11	toxic	toxic	ADJ
bracis-33560	230	12	words	word	NOUN
bracis-33560	230	13	,	,	PUNCT
bracis-33560	230	14	decreasing	decrease	VERB
bracis-33560	230	15	to	to	PART
bracis-33560	230	16	close	close	VERB
bracis-33560	230	17	to	to	ADP
bracis-33560	230	18	0	0	NUM
bracis-33560	230	19	(	(	PUNCT
bracis-33560	230	20	no	no	DET
bracis-33560	230	21	potentially	potentially	ADV
bracis-33560	230	22	toxic	toxic	ADJ
bracis-33560	230	23	words	word	NOUN
bracis-33560	230	24	)	)	PUNCT
bracis-33560	230	25	.	.	PUNCT
bracis-33560	231	1	the	the	DET
bracis-33560	231	2	results	result	NOUN
bracis-33560	231	3	show	show	VERB
bracis-33560	231	4	that	that	SCONJ
bracis-33560	231	5	we	we	PRON
bracis-33560	231	6	achieved	achieve	VERB
bracis-33560	231	7	this	this	DET
bracis-33560	231	8	goal	goal	NOUN
bracis-33560	231	9	in	in	ADP
bracis-33560	231	10	almost	almost	ADV
bracis-33560	231	11	all	all	DET
bracis-33560	231	12	the	the	DET
bracis-33560	231	13	texts	text	NOUN
bracis-33560	231	14	in	in	ADP
bracis-33560	231	15	the	the	DET
bracis-33560	231	16	corpus	corpus	NOUN
bracis-33560	231	17	,	,	PUNCT
bracis-33560	231	18	except	except	SCONJ
bracis-33560	231	19	for	for	ADP
bracis-33560	231	20	some	some	DET
bracis-33560	231	21	outliers	outlier	NOUN
bracis-33560	231	22	that	that	PRON
bracis-33560	231	23	do	do	AUX
bracis-33560	231	24	not	not	PART
bracis-33560	231	25	exceed	exceed	VERB
bracis-33560	231	26	0.2	0.2	NUM
bracis-33560	231	27	%	%	NOUN
bracis-33560	231	28	of	of	ADP
bracis-33560	231	29	toxic	toxic	ADJ
bracis-33560	231	30	content	content	NOUN
bracis-33560	231	31	.	.	PUNCT
bracis-33560	232	1	5.5	5.5	NUM
bracis-33560	232	2	bias	bias	NOUN
bracis-33560	232	3	we	we	PRON
bracis-33560	232	4	performed	perform	VERB
bracis-33560	232	5	a	a	DET
bracis-33560	232	6	word	word	NOUN
bracis-33560	232	7	co	co	NOUN
bracis-33560	232	8	-	-	NOUN
bracis-33560	232	9	occurrence	occurrence	ADJ
bracis-33560	232	10	analysis	analysis	NOUN
bracis-33560	232	11	to	to	PART
bracis-33560	232	12	identify	identify	VERB
bracis-33560	232	13	biases	bias	NOUN
bracis-33560	232	14	in	in	ADP
bracis-33560	232	15	our	our	PRON
bracis-33560	232	16	corpus	corpus	NOUN
bracis-33560	232	17	.	.	PUNCT
bracis-33560	233	1	this	this	DET
bracis-33560	233	2	technique	technique	NOUN
bracis-33560	233	3	has	have	AUX
bracis-33560	233	4	proven	prove	VERB
bracis-33560	233	5	effective	effective	ADJ
bracis-33560	233	6	in	in	ADP
bracis-33560	233	7	demonstrating	demonstrate	VERB
bracis-33560	233	8	stereotypical	stereotypical	ADJ
bracis-33560	233	9	treatment	treatment	NOUN
bracis-33560	233	10	of	of	ADP
bracis-33560	233	11	a	a	DET
bracis-33560	233	12	particular	particular	ADJ
bracis-33560	233	13	social	social	ADJ
bracis-33560	233	14	group	group	NOUN
bracis-33560	234	1	[	[	X
bracis-33560	234	2	13	13	NUM
bracis-33560	234	3	]	]	PUNCT
bracis-33560	234	4	,	,	PUNCT
bracis-33560	234	5	which	which	PRON
bracis-33560	234	6	we	we	PRON
bracis-33560	234	7	do	do	AUX
bracis-33560	234	8	not	not	PART
bracis-33560	234	9	desire	desire	VERB
bracis-33560	234	10	.	.	PUNCT
bracis-33560	235	1	it	it	PRON
bracis-33560	235	2	is	be	AUX
bracis-33560	235	3	worth	worth	ADJ
bracis-33560	235	4	noting	note	VERB
bracis-33560	235	5	that	that	SCONJ
bracis-33560	235	6	this	this	DET
bracis-33560	235	7	method	method	NOUN
bracis-33560	235	8	is	be	AUX
bracis-33560	235	9	one	one	NUM
bracis-33560	235	10	of	of	ADP
bracis-33560	235	11	several	several	ADJ
bracis-33560	235	12	mitigation	mitigation	NOUN
bracis-33560	235	13	approaches	approach	NOUN
bracis-33560	235	14	,	,	PUNCT
bracis-33560	235	15	and	and	CCONJ
bracis-33560	235	16	the	the	DET
bracis-33560	235	17	corpus	corpus	NOUN
bracis-33560	235	18	may	may	AUX
bracis-33560	235	19	still	still	ADV
bracis-33560	235	20	exhibit	exhibit	VERB
bracis-33560	235	21	bias	bias	NOUN
bracis-33560	235	22	.	.	PUNCT
bracis-33560	236	1	we	we	PRON
bracis-33560	236	2	analyzed	analyze	VERB
bracis-33560	236	3	three	three	NUM
bracis-33560	236	4	groups	group	NOUN
bracis-33560	236	5	of	of	ADP
bracis-33560	236	6	biases	bias	NOUN
bracis-33560	236	7	as	as	SCONJ
bracis-33560	236	8	shown	show	VERB
bracis-33560	236	9	in	in	ADP
bracis-33560	236	10	table	table	NOUN
bracis-33560	236	11	 	 	SPACE
bracis-33560	236	12	3	3	NUM
bracis-33560	236	13	,	,	PUNCT
bracis-33560	236	14	they	they	PRON
bracis-33560	236	15	are	be	AUX
bracis-33560	236	16	:	:	PUNCT
bracis-33560	236	17	(	(	PUNCT
bracis-33560	236	18	i	i	NOUN
bracis-33560	236	19	)	)	PUNCT
bracis-33560	236	20	gender	gender	NOUN
bracis-33560	236	21	,	,	PUNCT
bracis-33560	236	22	(	(	PUNCT
bracis-33560	236	23	ii	ii	NOUN
bracis-33560	236	24	)	)	PUNCT
bracis-33560	236	25	religion	religion	NOUN
bracis-33560	236	26	,	,	PUNCT
bracis-33560	236	27	and	and	CCONJ
bracis-33560	236	28	(	(	PUNCT
bracis-33560	236	29	iii	iii	NOUN
bracis-33560	236	30	)	)	PUNCT
bracis-33560	236	31	race	race	NOUN
bracis-33560	236	32	.	.	PUNCT
bracis-33560	237	1	we	we	PRON
bracis-33560	237	2	selected	select	VERB
bracis-33560	237	3	different	different	ADJ
bracis-33560	237	4	words	word	NOUN
bracis-33560	237	5	for	for	ADP
bracis-33560	237	6	each	each	DET
bracis-33560	237	7	group	group	NOUN
bracis-33560	237	8	and	and	CCONJ
bracis-33560	237	9	analyzed	analyze	VERB
bracis-33560	237	10	the	the	DET
bracis-33560	237	11	context	context	NOUN
bracis-33560	237	12	in	in	ADP
bracis-33560	237	13	which	which	PRON
bracis-33560	237	14	these	these	DET
bracis-33560	237	15	words	word	NOUN
bracis-33560	237	16	were	be	AUX
bracis-33560	237	17	inserted	insert	VERB
bracis-33560	237	18	.	.	PUNCT
bracis-33560	238	1	we	we	PRON
bracis-33560	238	2	have	have	AUX
bracis-33560	238	3	chosen	choose	VERB
bracis-33560	238	4	representative	representative	ADJ
bracis-33560	238	5	terms	term	NOUN
bracis-33560	238	6	indicative	indicative	ADJ
bracis-33560	238	7	of	of	ADP
bracis-33560	238	8	social	social	ADJ
bracis-33560	238	9	groups	group	NOUN
bracis-33560	238	10	and	and	CCONJ
bracis-33560	238	11	conducted	conduct	VERB
bracis-33560	238	12	an	an	DET
bracis-33560	238	13	analysis	analysis	NOUN
bracis-33560	238	14	focusing	focus	VERB
bracis-33560	238	15	on	on	ADP
bracis-33560	238	16	those	those	PRON
bracis-33560	238	17	with	with	ADP
bracis-33560	238	18	the	the	DET
bracis-33560	238	19	highest	high	ADJ
bracis-33560	238	20	co	co	ADJ
bracis-33560	238	21	-	-	ADJ
bracis-33560	238	22	occurrence	occurrence	ADJ
bracis-33560	238	23	frequencies	frequency	NOUN
bracis-33560	238	24	.	.	PUNCT
bracis-33560	239	1	we	we	PRON
bracis-33560	239	2	isolate	isolate	VERB
bracis-33560	239	3	the	the	DET
bracis-33560	239	4	terms	term	NOUN
bracis-33560	239	5	“	"	PUNCT
bracis-33560	239	6	man	man	NOUN
bracis-33560	239	7	”	"	PUNCT
bracis-33560	239	8	and	and	CCONJ
bracis-33560	239	9	“	"	PUNCT
bracis-33560	239	10	woman	woman	NOUN
bracis-33560	239	11	”	"	PUNCT
bracis-33560	239	12	to	to	PART
bracis-33560	239	13	evaluate	evaluate	VERB
bracis-33560	239	14	gender	gender	NOUN
bracis-33560	239	15	bias	bias	NOUN
bracis-33560	239	16	.	.	PUNCT
bracis-33560	240	1	we	we	PRON
bracis-33560	240	2	selected	select	VERB
bracis-33560	240	3	“	"	PUNCT
bracis-33560	240	4	atheist	atheist	NOUN
bracis-33560	240	5	”	"	PUNCT
bracis-33560	240	6	,	,	PUNCT
bracis-33560	240	7	“	"	PUNCT
bracis-33560	240	8	christian	christian	PROPN
bracis-33560	240	9	”	"	PUNCT
bracis-33560	240	10	,	,	PUNCT
bracis-33560	240	11	“	"	PUNCT
bracis-33560	240	12	buddhist	buddhist	ADJ
bracis-33560	240	13	”	"	PUNCT
bracis-33560	240	14	,	,	PUNCT
bracis-33560	240	15	“	"	PUNCT
bracis-33560	240	16	evangelical	evangelical	ADJ
bracis-33560	240	17	”	"	PUNCT
bracis-33560	240	18	,	,	PUNCT
bracis-33560	240	19	“	"	PUNCT
bracis-33560	240	20	jewish	jewish	ADJ
bracis-33560	240	21	”	"	PUNCT
bracis-33560	240	22	,	,	PUNCT
bracis-33560	240	23	“	"	PUNCT
bracis-33560	240	24	muslim	muslim	ADJ
bracis-33560	240	25	”	"	PUNCT
bracis-33560	240	26	and	and	CCONJ
bracis-33560	240	27	“	"	PUNCT
bracis-33560	240	28	umbandist	umbandist	NOUN
bracis-33560	240	29	”	"	PUNCT
bracis-33560	240	30	for	for	ADP
bracis-33560	240	31	religious	religious	ADJ
bracis-33560	240	32	bias	bias	NOUN
bracis-33560	240	33	.	.	PUNCT
bracis-33560	241	1	finally	finally	ADV
bracis-33560	241	2	,	,	PUNCT
bracis-33560	241	3	we	we	PRON
bracis-33560	241	4	highlighted	highlight	VERB
bracis-33560	241	5	the	the	DET
bracis-33560	241	6	words	word	NOUN
bracis-33560	241	7	“	"	PUNCT
bracis-33560	241	8	white	white	ADJ
bracis-33560	241	9	”	"	PUNCT
bracis-33560	241	10	,	,	PUNCT
bracis-33560	241	11	“	"	PUNCT
bracis-33560	241	12	black	black	ADJ
bracis-33560	241	13	”	"	PUNCT
bracis-33560	241	14	,	,	PUNCT
bracis-33560	241	15	“	"	PUNCT
bracis-33560	241	16	asian	asian	ADJ
bracis-33560	241	17	”	"	PUNCT
bracis-33560	241	18	and	and	CCONJ
bracis-33560	241	19	“	"	PUNCT
bracis-33560	241	20	hispanic	hispanic	ADJ
bracis-33560	241	21	”	"	PUNCT
bracis-33560	241	22	for	for	ADP
bracis-33560	241	23	racial	racial	ADJ
bracis-33560	241	24	bias	bias	NOUN
bracis-33560	241	25	.	.	PUNCT
bracis-33560	242	1	the	the	DET
bracis-33560	242	2	tables	table	NOUN
bracis-33560	242	3	 	 	SPACE
bracis-33560	242	4	3	3	NUM
bracis-33560	242	5	respectively	respectively	ADV
bracis-33560	242	6	represent	represent	VERB
bracis-33560	242	7	the	the	DET
bracis-33560	242	8	results	result	NOUN
bracis-33560	242	9	for	for	ADP
bracis-33560	242	10	gender	gender	NOUN
bracis-33560	242	11	,	,	PUNCT
bracis-33560	242	12	religious	religious	ADJ
bracis-33560	242	13	,	,	PUNCT
bracis-33560	242	14	and	and	CCONJ
bracis-33560	242	15	racial	racial	ADJ
bracis-33560	242	16	bias	bias	NOUN
bracis-33560	242	17	.	.	PUNCT
bracis-33560	243	1	table	table	NOUN
bracis-33560	243	2	3	3	NUM
bracis-33560	243	3	.	.	PUNCT
bracis-33560	243	4	related	relate	VERB
bracis-33560	243	5	word	word	NOUN
bracis-33560	243	6	co-occurrence.full	co-occurrence.full	NOUN
bracis-33560	243	7	size	size	NOUN
bracis-33560	243	8	table	table	NOUN
bracis-33560	243	9	no	no	DET
bracis-33560	243	10	stereotypical	stereotypical	ADJ
bracis-33560	243	11	treatment	treatment	NOUN
bracis-33560	243	12	or	or	CCONJ
bracis-33560	243	13	hazardous	hazardous	ADJ
bracis-33560	243	14	behaviors	behavior	NOUN
bracis-33560	243	15	described	describe	VERB
bracis-33560	243	16	in	in	ADP
bracis-33560	243	17	the	the	DET
bracis-33560	243	18	literature	literature	NOUN
bracis-33560	243	19	exist	exist	VERB
bracis-33560	243	20	in	in	ADP
bracis-33560	243	21	the	the	DET
bracis-33560	243	22	analyzed	analyze	VERB
bracis-33560	243	23	groups	group	NOUN
bracis-33560	243	24	,	,	PUNCT
bracis-33560	243	25	such	such	ADJ
bracis-33560	243	26	as	as	ADP
bracis-33560	243	27	associations	association	NOUN
bracis-33560	243	28	with	with	ADP
bracis-33560	243	29	crime	crime	NOUN
bracis-33560	243	30	,	,	PUNCT
bracis-33560	243	31	income	income	NOUN
bracis-33560	243	32	,	,	PUNCT
bracis-33560	243	33	and	and	CCONJ
bracis-33560	243	34	others	other	NOUN
bracis-33560	243	35	 	 	SPACE
bracis-33560	244	1	[	[	X
bracis-33560	244	2	16	16	NUM
bracis-33560	244	3	]	]	PUNCT
bracis-33560	244	4	.	.	PUNCT
bracis-33560	245	1	furthermore	furthermore	ADV
bracis-33560	245	2	,	,	PUNCT
bracis-33560	245	3	the	the	DET
bracis-33560	245	4	words	word	NOUN
bracis-33560	245	5	selected	select	VERB
bracis-33560	245	6	among	among	ADP
bracis-33560	245	7	the	the	DET
bracis-33560	245	8	various	various	ADJ
bracis-33560	245	9	social	social	ADJ
bracis-33560	245	10	groups	group	NOUN
bracis-33560	245	11	are	be	AUX
bracis-33560	245	12	very	very	ADV
bracis-33560	245	13	similar	similar	ADJ
bracis-33560	245	14	,	,	PUNCT
bracis-33560	245	15	which	which	PRON
bracis-33560	245	16	suggests	suggest	VERB
bracis-33560	245	17	a	a	DET
bracis-33560	245	18	more	more	ADV
bracis-33560	245	19	equanimous	equanimous	ADJ
bracis-33560	245	20	treatment	treatment	NOUN
bracis-33560	245	21	in	in	ADP
bracis-33560	245	22	our	our	PRON
bracis-33560	245	23	proposed	propose	VERB
bracis-33560	245	24	corpus	corpus	NOUN
bracis-33560	245	25	.	.	PUNCT
bracis-33560	246	1	6	6	NUM
bracis-33560	246	2	conclusion	conclusion	NOUN
bracis-33560	246	3	recent	recent	ADJ
bracis-33560	246	4	studies	study	NOUN
bracis-33560	246	5	have	have	AUX
bracis-33560	246	6	shown	show	VERB
bracis-33560	246	7	significant	significant	ADJ
bracis-33560	246	8	improvements	improvement	NOUN
bracis-33560	246	9	in	in	ADP
bracis-33560	246	10	the	the	DET
bracis-33560	246	11	performance	performance	NOUN
bracis-33560	246	12	of	of	ADP
bracis-33560	246	13	language	language	NOUN
bracis-33560	246	14	models	model	NOUN
bracis-33560	246	15	trained	train	VERB
bracis-33560	246	16	on	on	ADP
bracis-33560	246	17	large	large	ADJ
bracis-33560	246	18	corpora	corpora	NOUN
bracis-33560	247	1	[	[	X
bracis-33560	247	2	8	8	NUM
bracis-33560	247	3	,	,	PUNCT
bracis-33560	247	4	14	14	NUM
bracis-33560	247	5	]	]	PUNCT
bracis-33560	247	6	.	.	PUNCT
bracis-33560	248	1	consequently	consequently	ADV
bracis-33560	248	2	,	,	PUNCT
bracis-33560	248	3	the	the	DET
bracis-33560	248	4	interest	interest	NOUN
bracis-33560	248	5	in	in	ADP
bracis-33560	248	6	creating	create	VERB
bracis-33560	248	7	large	large	ADJ
bracis-33560	248	8	datasets	dataset	NOUN
bracis-33560	248	9	has	have	AUX
bracis-33560	248	10	grown	grow	VERB
bracis-33560	248	11	.	.	PUNCT
bracis-33560	249	1	most	most	ADJ
bracis-33560	249	2	existing	exist	VERB
bracis-33560	249	3	research	research	NOUN
bracis-33560	249	4	focuses	focus	VERB
bracis-33560	249	5	on	on	ADP
bracis-33560	249	6	high	high	ADJ
bracis-33560	249	7	-	-	PUNCT
bracis-33560	249	8	resource	resource	NOUN
bracis-33560	249	9	languages	language	NOUN
bracis-33560	249	10	like	like	ADP
bracis-33560	249	11	english	english	PROPN
bracis-33560	249	12	and	and	CCONJ
bracis-33560	249	13	chinese	chinese	PROPN
bracis-33560	249	14	,	,	PUNCT
bracis-33560	249	15	with	with	ADP
bracis-33560	249	16	considerable	considerable	ADJ
bracis-33560	249	17	efforts	effort	NOUN
bracis-33560	249	18	made	make	VERB
bracis-33560	249	19	to	to	PART
bracis-33560	249	20	develop	develop	VERB
bracis-33560	249	21	multilingual	multilingual	ADJ
bracis-33560	249	22	corpora	corpus	NOUN
bracis-33560	249	23	.	.	PUNCT
bracis-33560	250	1	however	however	ADV
bracis-33560	250	2	,	,	PUNCT
bracis-33560	250	3	there	there	PRON
bracis-33560	250	4	is	be	VERB
bracis-33560	250	5	a	a	DET
bracis-33560	250	6	pressing	press	VERB
bracis-33560	250	7	need	need	NOUN
bracis-33560	250	8	to	to	PART
bracis-33560	250	9	develop	develop	VERB
bracis-33560	250	10	large	large	ADJ
bracis-33560	250	11	datasets	dataset	NOUN
bracis-33560	250	12	for	for	ADP
bracis-33560	250	13	lower	low	ADJ
bracis-33560	250	14	-	-	PUNCT
bracis-33560	250	15	resource	resource	NOUN
bracis-33560	250	16	languages	language	NOUN
bracis-33560	250	17	.	.	PUNCT
bracis-33560	251	1	this	this	DET
bracis-33560	251	2	work	work	NOUN
bracis-33560	251	3	aims	aim	VERB
bracis-33560	251	4	to	to	PART
bracis-33560	251	5	address	address	VERB
bracis-33560	251	6	this	this	DET
bracis-33560	251	7	gap	gap	NOUN
bracis-33560	251	8	by	by	ADP
bracis-33560	251	9	developing	develop	VERB
bracis-33560	251	10	language	language	NOUN
bracis-33560	251	11	models	model	NOUN
bracis-33560	251	12	for	for	ADP
bracis-33560	251	13	lower	low	ADJ
bracis-33560	251	14	-	-	PUNCT
bracis-33560	251	15	resource	resource	NOUN
bracis-33560	251	16	languages	language	NOUN
bracis-33560	251	17	,	,	PUNCT
bracis-33560	251	18	specifically	specifically	ADV
bracis-33560	251	19	portuguese	portuguese	ADJ
bracis-33560	251	20	.	.	PUNCT
bracis-33560	252	1	we	we	PRON
bracis-33560	252	2	created	create	VERB
bracis-33560	252	3	the	the	DET
bracis-33560	252	4	largest	large	ADJ
bracis-33560	252	5	curated	curate	VERB
bracis-33560	252	6	corpora	corpora	NOUN
bracis-33560	252	7	for	for	ADP
bracis-33560	252	8	training	training	NOUN
bracis-33560	252	9	or	or	CCONJ
bracis-33560	252	10	pre	pre	ADJ
bracis-33560	252	11	-	-	ADJ
bracis-33560	252	12	training	training	ADJ
bracis-33560	252	13	portuguese	portuguese	ADJ
bracis-33560	252	14	language	language	NOUN
bracis-33560	252	15	models	model	NOUN
bracis-33560	252	16	.	.	PUNCT
bracis-33560	253	1	to	to	PART
bracis-33560	253	2	achieve	achieve	VERB
bracis-33560	253	3	this	this	PRON
bracis-33560	253	4	,	,	PUNCT
bracis-33560	253	5	we	we	PRON
bracis-33560	253	6	implemented	implement	VERB
bracis-33560	253	7	a	a	DET
bracis-33560	253	8	double	double	ADJ
bracis-33560	253	9	-	-	PUNCT
bracis-33560	253	10	pipeline	pipeline	NOUN
bracis-33560	253	11	process	process	NOUN
bracis-33560	253	12	to	to	PART
bracis-33560	253	13	extract	extract	VERB
bracis-33560	253	14	data	datum	NOUN
bracis-33560	253	15	while	while	SCONJ
bracis-33560	253	16	ensuring	ensure	VERB
bracis-33560	253	17	content	content	NOUN
bracis-33560	253	18	safety	safety	NOUN
bracis-33560	253	19	.	.	PUNCT
bracis-33560	254	1	the	the	DET
bracis-33560	254	2	process	process	NOUN
bracis-33560	254	3	includes	include	VERB
bracis-33560	254	4	downloading	download	VERB
bracis-33560	254	5	,	,	PUNCT
bracis-33560	254	6	text	text	NOUN
bracis-33560	254	7	extraction	extraction	NOUN
bracis-33560	254	8	,	,	PUNCT
bracis-33560	254	9	language	language	NOUN
bracis-33560	254	10	identification	identification	NOUN
bracis-33560	254	11	,	,	PUNCT
bracis-33560	254	12	deduplication	deduplication	NOUN
bracis-33560	254	13	,	,	PUNCT
bracis-33560	254	14	quality	quality	NOUN
bracis-33560	254	15	filtering	filtering	NOUN
bracis-33560	254	16	,	,	PUNCT
bracis-33560	254	17	filtering	filter	VERB
bracis-33560	254	18	for	for	ADP
bracis-33560	254	19	sexual	sexual	ADJ
bracis-33560	254	20	content	content	NOUN
bracis-33560	254	21	and	and	CCONJ
bracis-33560	254	22	toxicity	toxicity	NOUN
bracis-33560	254	23	,	,	PUNCT
bracis-33560	254	24	bias	bias	NOUN
bracis-33560	254	25	filtering	filtering	NOUN
bracis-33560	254	26	,	,	PUNCT
bracis-33560	254	27	categorization	categorization	NOUN
bracis-33560	254	28	,	,	PUNCT
bracis-33560	254	29	and	and	CCONJ
bracis-33560	254	30	storage	storage	NOUN
bracis-33560	254	31	.	.	PUNCT
bracis-33560	255	1	this	this	DET
bracis-33560	255	2	effort	effort	NOUN
bracis-33560	255	3	involved	involve	VERB
bracis-33560	255	4	collecting	collect	VERB
bracis-33560	255	5	terabytes	terabyte	NOUN
bracis-33560	255	6	of	of	ADP
bracis-33560	255	7	data	datum	NOUN
bracis-33560	255	8	,	,	PUNCT
bracis-33560	255	9	resulting	result	VERB
bracis-33560	255	10	in	in	ADP
bracis-33560	255	11	a	a	DET
bracis-33560	255	12	curated	curate	VERB
bracis-33560	255	13	dataset	dataset	NOUN
bracis-33560	255	14	of	of	ADP
bracis-33560	255	15	approximately	approximately	ADV
bracis-33560	255	16	100	100	NUM
bracis-33560	255	17	gb	gb	NOUN
bracis-33560	255	18	for	for	ADP
bracis-33560	255	19	aroeira	aroeira	PROPN
bracis-33560	255	20	’s	’s	PART
bracis-33560	255	21	construction	construction	NOUN
bracis-33560	255	22	.	.	PUNCT
bracis-33560	256	1	our	our	PRON
bracis-33560	256	2	results	result	NOUN
bracis-33560	256	3	demonstrate	demonstrate	VERB
bracis-33560	256	4	that	that	SCONJ
bracis-33560	256	5	our	our	PRON
bracis-33560	256	6	corpus	corpus	NOUN
bracis-33560	256	7	fulfills	fulfill	VERB
bracis-33560	256	8	the	the	DET
bracis-33560	256	9	requirements	requirement	NOUN
bracis-33560	256	10	of	of	ADP
bracis-33560	256	11	corpus	corpus	NOUN
bracis-33560	256	12	size	size	NOUN
bracis-33560	256	13	,	,	PUNCT
bracis-33560	256	14	knowledge	knowledge	NOUN
bracis-33560	256	15	domains	domain	NOUN
bracis-33560	256	16	,	,	PUNCT
bracis-33560	256	17	document	document	NOUN
bracis-33560	256	18	distribution	distribution	NOUN
bracis-33560	256	19	over	over	ADP
bracis-33560	256	20	time	time	NOUN
bracis-33560	256	21	,	,	PUNCT
bracis-33560	256	22	quality	quality	NOUN
bracis-33560	256	23	indicators	indicator	NOUN
bracis-33560	256	24	,	,	PUNCT
bracis-33560	256	25	and	and	CCONJ
bracis-33560	256	26	bias	bias	NOUN
bracis-33560	256	27	mitigation	mitigation	NOUN
bracis-33560	256	28	.	.	PUNCT
bracis-33560	257	1	we	we	PRON
bracis-33560	257	2	conducted	conduct	VERB
bracis-33560	257	3	statistical	statistical	ADJ
bracis-33560	257	4	analyses	analysis	NOUN
bracis-33560	257	5	of	of	ADP
bracis-33560	257	6	the	the	DET
bracis-33560	257	7	corpus	corpus	NOUN
bracis-33560	257	8	to	to	PART
bracis-33560	257	9	better	well	ADV
bracis-33560	257	10	comprehend	comprehend	VERB
bracis-33560	257	11	the	the	DET
bracis-33560	257	12	size	size	NOUN
bracis-33560	257	13	of	of	ADP
bracis-33560	257	14	the	the	DET
bracis-33560	257	15	collected	collect	VERB
bracis-33560	257	16	documents	document	NOUN
bracis-33560	257	17	in	in	ADP
bracis-33560	257	18	terms	term	NOUN
bracis-33560	257	19	of	of	ADP
bracis-33560	257	20	the	the	DET
bracis-33560	257	21	number	number	NOUN
bracis-33560	257	22	of	of	ADP
bracis-33560	257	23	tokens	token	NOUN
bracis-33560	257	24	,	,	PUNCT
bracis-33560	257	25	words	word	NOUN
bracis-33560	257	26	,	,	PUNCT
bracis-33560	257	27	and	and	CCONJ
bracis-33560	257	28	sentences	sentence	NOUN
bracis-33560	257	29	.	.	PUNCT
bracis-33560	258	1	additionally	additionally	ADV
bracis-33560	258	2	,	,	PUNCT
bracis-33560	258	3	we	we	PRON
bracis-33560	258	4	analyzed	analyze	VERB
bracis-33560	258	5	bias	bias	NOUN
bracis-33560	258	6	to	to	PART
bracis-33560	258	7	recognize	recognize	VERB
bracis-33560	258	8	potential	potential	ADJ
bracis-33560	258	9	harms	harm	NOUN
bracis-33560	258	10	in	in	ADP
bracis-33560	258	11	the	the	DET
bracis-33560	258	12	created	create	VERB
bracis-33560	258	13	corpus	corpus	NOUN
bracis-33560	258	14	,	,	PUNCT
bracis-33560	258	15	a	a	DET
bracis-33560	258	16	distinguishing	distinguish	VERB
bracis-33560	258	17	factor	factor	NOUN
bracis-33560	258	18	in	in	ADP
bracis-33560	258	19	building	building	NOUN
bracis-33560	258	20	models	model	NOUN
bracis-33560	258	21	free	free	ADJ
bracis-33560	258	22	from	from	ADP
bracis-33560	258	23	social	social	ADJ
bracis-33560	258	24	biases	bias	NOUN
bracis-33560	258	25	.	.	PUNCT
bracis-33560	259	1	our	our	PRON
bracis-33560	259	2	findings	finding	NOUN
bracis-33560	259	3	conclude	conclude	VERB
bracis-33560	259	4	that	that	SCONJ
bracis-33560	259	5	the	the	DET
bracis-33560	259	6	corpus	corpus	NOUN
bracis-33560	259	7	created	create	VERB
bracis-33560	259	8	is	be	AUX
bracis-33560	259	9	of	of	ADP
bracis-33560	259	10	high	high	ADJ
bracis-33560	259	11	quality	quality	NOUN
bracis-33560	259	12	and	and	CCONJ
bracis-33560	259	13	diversity	diversity	NOUN
bracis-33560	259	14	,	,	PUNCT
bracis-33560	259	15	with	with	ADP
bracis-33560	259	16	minimal	minimal	ADJ
bracis-33560	259	17	bias	bias	NOUN
bracis-33560	259	18	.	.	PUNCT
bracis-33560	260	1	we	we	PRON
bracis-33560	260	2	are	be	AUX
bracis-33560	260	3	eager	eager	ADJ
bracis-33560	260	4	to	to	PART
bracis-33560	260	5	advance	advance	VERB
bracis-33560	260	6	our	our	PRON
bracis-33560	260	7	work	work	NOUN
bracis-33560	260	8	by	by	ADP
bracis-33560	260	9	training	training	NOUN
bracis-33560	260	10	models	model	NOUN
bracis-33560	260	11	with	with	ADP
bracis-33560	260	12	encoder	encoder	NOUN
bracis-33560	260	13	architectures	architecture	NOUN
bracis-33560	260	14	,	,	PUNCT
bracis-33560	260	15	such	such	ADJ
bracis-33560	260	16	as	as	ADP
bracis-33560	260	17	bert	bert	PROPN
bracis-33560	260	18	models	model	NOUN
bracis-33560	260	19	[	[	X
bracis-33560	260	20	21	21	NUM
bracis-33560	260	21	]	]	PUNCT
bracis-33560	260	22	.	.	PUNCT
bracis-33560	261	1	furthermore	furthermore	ADV
bracis-33560	261	2	,	,	PUNCT
bracis-33560	261	3	we	we	PRON
bracis-33560	261	4	plan	plan	VERB
bracis-33560	261	5	to	to	PART
bracis-33560	261	6	pre	pre	VERB
bracis-33560	261	7	-	-	ADJ
bracis-33560	261	8	train	train	ADJ
bracis-33560	261	9	language	language	NOUN
bracis-33560	261	10	models	model	NOUN
bracis-33560	261	11	with	with	ADP
bracis-33560	261	12	aroeira	aroeira	PROPN
bracis-33560	261	13	to	to	PART
bracis-33560	261	14	obtain	obtain	VERB
bracis-33560	261	15	higher	high	ADJ
bracis-33560	261	16	-	-	PUNCT
bracis-33560	261	17	quality	quality	NOUN
bracis-33560	261	18	models	model	NOUN
bracis-33560	261	19	in	in	ADP
bracis-33560	261	20	portuguese	portuguese	PROPN
bracis-33560	261	21	.	.	PUNCT
bracis-33560	262	1	we	we	PRON
bracis-33560	262	2	intend	intend	VERB
bracis-33560	262	3	to	to	PART
bracis-33560	262	4	research	research	VERB
bracis-33560	262	5	existent	existent	ADJ
bracis-33560	262	6	instruction	instruction	NOUN
bracis-33560	262	7	and	and	CCONJ
bracis-33560	262	8	evaluation	evaluation	NOUN
bracis-33560	262	9	datasets	dataset	NOUN
bracis-33560	262	10	.	.	PUNCT
bracis-33560	263	1	finally	finally	ADV
bracis-33560	263	2	,	,	PUNCT
bracis-33560	263	3	conducting	conduct	VERB
bracis-33560	263	4	comparative	comparative	ADJ
bracis-33560	263	5	bias	bias	NOUN
bracis-33560	263	6	analyses	analysis	NOUN
bracis-33560	263	7	on	on	ADP
bracis-33560	263	8	models	model	NOUN
bracis-33560	263	9	trained	train	VERB
bracis-33560	263	10	with	with	ADP
bracis-33560	263	11	aroeira	aroeira	PROPN
bracis-33560	263	12	relative	relative	ADJ
bracis-33560	263	13	to	to	ADP
bracis-33560	263	14	other	other	ADJ
bracis-33560	263	15	available	available	ADJ
bracis-33560	263	16	corpora	corpora	PROPN
bracis-33560	264	1	[	[	X
bracis-33560	264	2	13	13	NUM
bracis-33560	264	3	]	]	PUNCT
bracis-33560	264	4	will	will	AUX
bracis-33560	264	5	be	be	AUX
bracis-33560	264	6	highly	highly	ADV
bracis-33560	264	7	valuable	valuable	ADJ
bracis-33560	264	8	.	.	PUNCT
bracis-33560	265	1	notes	note	VERB
bracis-33560	265	2	1.available	1.available	NUM
bracis-33560	265	3	at	at	ADP
bracis-33560	265	4	:	:	PUNCT
bracis-33560	265	5	https://commoncrawl.org/.	https://commoncrawl.org/.	PROPN
bracis-33560	265	6	2.available	2.available	NUM
bracis-33560	265	7	at	at	ADP
bracis-33560	265	8	:	:	PUNCT
bracis-33560	265	9	https://fasttext.cc/.	https://fasttext.cc/.	PROPN
bracis-33560	265	10	3.https://dsi.ut-capitole.fr/blacklists/index_en.php	3.https://dsi.ut-capitole.fr/blacklists/index_en.php	NUM
bracis-33560	265	11	.	.	PUNCT
bracis-33560	266	1	4.https://github.com/ldnoobw/list-of-dirty-naughty-obscene-and-otherwise-bad-words/blob/master/pt	4.https://github.com/ldnoobw/list-of-dirty-naughty-obscene-and-otherwise-bad-words/blob/master/pt	NUM
bracis-33560	266	2	.	.	PUNCT
bracis-33560	267	1	5.https://github.com/dunossauro/chat-detox/blob/main/palavras.txt	5.https://github.com/dunossauro/chat-detox/blob/main/palavras.txt	NUM
bracis-33560	267	2	.	.	PUNCT
bracis-33560	268	1	references	reference	NOUN
bracis-33560	268	2	abadji	abadji	PROPN
bracis-33560	268	3	,	,	PUNCT
bracis-33560	268	4	j.	j.	PROPN
bracis-33560	268	5	,	,	PUNCT
bracis-33560	268	6	ortiz	ortiz	PROPN
bracis-33560	268	7	 	 	SPACE
bracis-33560	268	8	suárez	suárez	PROPN
bracis-33560	268	9	,	,	PUNCT
bracis-33560	268	10	p.j	p.j	PROPN
bracis-33560	268	11	.	.	PROPN
bracis-33560	268	12	,	,	PUNCT
bracis-33560	268	13	romary	romary	PROPN
bracis-33560	268	14	,	,	PUNCT
bracis-33560	268	15	l.	l.	PROPN
bracis-33560	268	16	,	,	PUNCT
bracis-33560	268	17	sagot	sagot	PROPN
bracis-33560	268	18	,	,	PUNCT
bracis-33560	268	19	b.	b.	PROPN
bracis-33560	268	20	:	:	PUNCT
bracis-33560	269	1	ungoliant	ungoliant	PROPN
bracis-33560	269	2	:	:	PUNCT
bracis-33560	269	3	an	an	DET
bracis-33560	269	4	optimized	optimize	VERB
bracis-33560	269	5	pipeline	pipeline	NOUN
bracis-33560	269	6	for	for	ADP
bracis-33560	269	7	the	the	DET
bracis-33560	269	8	generation	generation	NOUN
bracis-33560	269	9	of	of	ADP
bracis-33560	269	10	a	a	DET
bracis-33560	269	11	very	very	ADV
bracis-33560	269	12	large	large	ADJ
bracis-33560	269	13	-	-	PUNCT
bracis-33560	269	14	scale	scale	NOUN
bracis-33560	269	15	multilingual	multilingual	ADJ
bracis-33560	269	16	web	web	NOUN
bracis-33560	269	17	corpus	corpus	NOUN
bracis-33560	269	18	(	(	PUNCT
bracis-33560	269	19	2021	2021	NUM
bracis-33560	269	20	)	)	PUNCT
bracis-33560	269	21	.	.	PUNCT
bracis-33560	270	1	https://doi.org/10.14618/ids-pub-10468	https://doi.org/10.14618/ids-pub-10468	PROPN
bracis-33560	270	2	abadji	abadji	PROPN
bracis-33560	270	3	,	,	PUNCT
bracis-33560	270	4	j.	j.	PROPN
bracis-33560	270	5	,	,	PUNCT
bracis-33560	270	6	suarez	suarez	PROPN
bracis-33560	270	7	,	,	PUNCT
bracis-33560	270	8	p.o	p.o	PROPN
bracis-33560	270	9	.	.	PROPN
bracis-33560	270	10	,	,	PUNCT
bracis-33560	270	11	romary	romary	PROPN
bracis-33560	270	12	,	,	PUNCT
bracis-33560	270	13	l.	l.	PROPN
bracis-33560	270	14	,	,	PUNCT
bracis-33560	270	15	sagot	sagot	PROPN
bracis-33560	270	16	,	,	PUNCT
bracis-33560	270	17	b.	b.	PROPN
bracis-33560	270	18	:	:	PUNCT
bracis-33560	270	19	towards	towards	ADP
bracis-33560	270	20	a	a	DET
bracis-33560	270	21	cleaner	clean	ADJ
bracis-33560	270	22	document	document	NOUN
bracis-33560	270	23	-	-	PUNCT
bracis-33560	270	24	oriented	orient	VERB
bracis-33560	270	25	multilingual	multilingual	NOUN
bracis-33560	270	26	crawled	crawl	VERB
bracis-33560	270	27	corpus	corpu	NOUN
bracis-33560	270	28	,	,	PUNCT
bracis-33560	270	29	january	january	PROPN
bracis-33560	270	30	2022	2022	NUM
bracis-33560	270	31	google	google	PROPN
bracis-33560	270	32	scholar	scholar	NOUN
bracis-33560	270	33	  	  	SPACE
bracis-33560	270	34	agrawal	agrawal	PROPN
bracis-33560	270	35	,	,	PUNCT
bracis-33560	270	36	a.	a.	NOUN
bracis-33560	270	37	,	,	PUNCT
bracis-33560	270	38	singh	singh	PROPN
bracis-33560	270	39	,	,	PUNCT
bracis-33560	270	40	s.	s.	PROPN
bracis-33560	270	41	:	:	PUNCT
bracis-33560	270	42	corpus	corpus	NOUN
bracis-33560	270	43	complexity	complexity	NOUN
bracis-33560	270	44	matters	matter	VERB
bracis-33560	270	45	in	in	ADP
bracis-33560	270	46	pretraining	pretraine	VERB
bracis-33560	270	47	language	language	NOUN
bracis-33560	270	48	models	model	NOUN
bracis-33560	270	49	,	,	PUNCT
bracis-33560	270	50	pp	pp	ADP
bracis-33560	270	51	.	.	PUNCT
bracis-33560	271	1	257–263	257–263	NUM
bracis-33560	271	2	,	,	PUNCT
bracis-33560	271	3	january	january	PROPN
bracis-33560	271	4	2023	2023	NUM
bracis-33560	271	5	.	.	PUNCT
bracis-33560	272	1	https://doi.org/10.18653/v1/2023.sustainlp-1.20	https://doi.org/10.18653/v1/2023.sustainlp-1.20	PROPN
bracis-33560	272	2	almeida	almeida	PROPN
bracis-33560	272	3	,	,	PUNCT
bracis-33560	272	4	t.s	t.s	PROPN
bracis-33560	272	5	.	.	PROPN
bracis-33560	272	6	,	,	PUNCT
bracis-33560	272	7	abonizio	abonizio	PROPN
bracis-33560	272	8	,	,	PUNCT
bracis-33560	272	9	h.	h.	PROPN
bracis-33560	272	10	,	,	PUNCT
bracis-33560	272	11	nogueira	nogueira	PROPN
bracis-33560	272	12	,	,	PUNCT
bracis-33560	272	13	r.	r.	PROPN
bracis-33560	272	14	,	,	PUNCT
bracis-33560	272	15	pires	pire	NOUN
bracis-33560	272	16	,	,	PUNCT
bracis-33560	272	17	r.	r.	PROPN
bracis-33560	272	18	:	:	PUNCT
bracis-33560	272	19	sabi\(\backslash	sabi\(\backslash	PROPN
bracis-33560	272	20	\)’a-2	\)’a-2	PROPN
bracis-33560	272	21	:	:	PUNCT
bracis-33560	272	22	a	a	DET
bracis-33560	272	23	new	new	ADJ
bracis-33560	272	24	generation	generation	NOUN
bracis-33560	272	25	of	of	ADP
bracis-33560	272	26	portuguese	portuguese	ADJ
bracis-33560	272	27	large	large	ADJ
bracis-33560	272	28	language	language	NOUN
bracis-33560	272	29	models	model	NOUN
bracis-33560	272	30	.	.	PUNCT
bracis-33560	273	1	arxiv	arxiv	PROPN
bracis-33560	273	2	preprint	preprint	VERB
bracis-33560	273	3	arxiv:2403.09887	arxiv:2403.09887	NUM
bracis-33560	273	4	(	(	PUNCT
bracis-33560	273	5	2024	2024	NUM
bracis-33560	273	6	)	)	PUNCT
bracis-33560	273	7	barbaresi	barbaresi	NOUN
bracis-33560	273	8	,	,	PUNCT
bracis-33560	273	9	a.	a.	NOUN
bracis-33560	273	10	:	:	PUNCT
bracis-33560	274	1	trafilatura	trafilatura	ADV
bracis-33560	274	2	:	:	PUNCT
bracis-33560	274	3	a	a	DET
bracis-33560	274	4	web	web	NOUN
bracis-33560	274	5	scraping	scrape	VERB
bracis-33560	274	6	library	library	NOUN
bracis-33560	274	7	and	and	CCONJ
bracis-33560	274	8	command	command	NOUN
bracis-33560	274	9	-	-	PUNCT
bracis-33560	274	10	line	line	NOUN
bracis-33560	274	11	tool	tool	NOUN
bracis-33560	274	12	for	for	ADP
bracis-33560	274	13	text	text	NOUN
bracis-33560	274	14	discovery	discovery	NOUN
bracis-33560	274	15	and	and	CCONJ
bracis-33560	274	16	extraction	extraction	NOUN
bracis-33560	274	17	.	.	PUNCT
bracis-33560	275	1	in	in	ADP
bracis-33560	275	2	:	:	PUNCT
bracis-33560	275	3	proceedings	proceeding	NOUN
bracis-33560	275	4	of	of	ADP
bracis-33560	275	5	the	the	DET
bracis-33560	275	6	joint	joint	ADJ
bracis-33560	275	7	conference	conference	NOUN
bracis-33560	275	8	of	of	ADP
bracis-33560	275	9	the	the	DET
bracis-33560	275	10	59th	59th	ADJ
bracis-33560	275	11	annual	annual	ADJ
bracis-33560	275	12	meeting	meeting	NOUN
bracis-33560	275	13	of	of	ADP
bracis-33560	275	14	the	the	DET
bracis-33560	275	15	association	association	NOUN
bracis-33560	275	16	for	for	ADP
bracis-33560	275	17	computational	computational	ADJ
bracis-33560	275	18	linguistics	linguistic	NOUN
bracis-33560	275	19	and	and	CCONJ
bracis-33560	275	20	the	the	DET
bracis-33560	275	21	11th	11th	ADJ
bracis-33560	275	22	international	international	ADJ
bracis-33560	275	23	joint	joint	ADJ
bracis-33560	275	24	conference	conference	NOUN
bracis-33560	275	25	on	on	ADP
bracis-33560	275	26	natural	natural	ADJ
bracis-33560	275	27	language	language	NOUN
bracis-33560	275	28	processing	processing	NOUN
bracis-33560	275	29	:	:	PUNCT
bracis-33560	275	30	system	system	NOUN
bracis-33560	275	31	demonstrations	demonstration	NOUN
bracis-33560	275	32	,	,	PUNCT
bracis-33560	275	33	pp	pp	ADJ
bracis-33560	275	34	.	.	PUNCT
bracis-33560	276	1	122–131	122–131	NUM
bracis-33560	276	2	.	.	X
bracis-33560	276	3	association	association	NOUN
bracis-33560	276	4	for	for	ADP
bracis-33560	276	5	computational	computational	ADJ
bracis-33560	276	6	linguistics	linguistic	NOUN
bracis-33560	276	7	(	(	PUNCT
bracis-33560	276	8	2021	2021	NUM
bracis-33560	276	9	)	)	PUNCT
bracis-33560	276	10	.	.	PUNCT
bracis-33560	277	1	https://aclanthology.org/2021.acl-demo.15	https://aclanthology.org/2021.acl-demo.15	PROPN
bracis-33560	277	2	blodgett	blodgett	PROPN
bracis-33560	277	3	,	,	PUNCT
bracis-33560	277	4	s.l	s.l	PROPN
bracis-33560	277	5	.	.	PROPN
bracis-33560	277	6	,	,	PUNCT
bracis-33560	277	7	barocas	barocas	PROPN
bracis-33560	277	8	,	,	PUNCT
bracis-33560	277	9	s.	s.	PROPN
bracis-33560	277	10	,	,	PUNCT
bracis-33560	277	11	daumé	daumé	PROPN
bracis-33560	277	12	 	 	SPACE
bracis-33560	277	13	iii	iii	PROPN
bracis-33560	277	14	,	,	PUNCT
bracis-33560	277	15	h.	h.	PROPN
bracis-33560	277	16	,	,	PUNCT
bracis-33560	277	17	wallach	wallach	PROPN
bracis-33560	277	18	,	,	PUNCT
bracis-33560	277	19	h.	h.	PROPN
bracis-33560	277	20	:	:	PUNCT
bracis-33560	277	21	language	language	NOUN
bracis-33560	277	22	(	(	PUNCT
bracis-33560	277	23	technology	technology	NOUN
bracis-33560	277	24	)	)	PUNCT
bracis-33560	277	25	is	be	AUX
bracis-33560	277	26	power	power	NOUN
bracis-33560	277	27	:	:	PUNCT
bracis-33560	277	28	a	a	DET
bracis-33560	277	29	critical	critical	ADJ
bracis-33560	277	30	survey	survey	NOUN
bracis-33560	277	31	of	of	ADP
bracis-33560	277	32	“	"	PUNCT
bracis-33560	277	33	bias	bias	NOUN
bracis-33560	277	34	”	"	PUNCT
bracis-33560	277	35	in	in	ADP
bracis-33560	277	36	nlp	nlp	NOUN
bracis-33560	277	37	.	.	PUNCT
bracis-33560	278	1	in	in	ADP
bracis-33560	278	2	:	:	PUNCT
bracis-33560	278	3	proceedings	proceeding	NOUN
bracis-33560	278	4	of	of	ADP
bracis-33560	278	5	the	the	DET
bracis-33560	278	6	58th	58th	ADJ
bracis-33560	278	7	annual	annual	ADJ
bracis-33560	278	8	meeting	meeting	NOUN
bracis-33560	278	9	of	of	ADP
bracis-33560	278	10	the	the	DET
bracis-33560	278	11	association	association	NOUN
bracis-33560	278	12	for	for	ADP
bracis-33560	278	13	computational	computational	ADJ
bracis-33560	278	14	linguistics	linguistic	NOUN
bracis-33560	278	15	.	.	PUNCT
bracis-33560	279	1	association	association	NOUN
bracis-33560	279	2	for	for	ADP
bracis-33560	279	3	computational	computational	ADJ
bracis-33560	279	4	linguistics	linguistic	NOUN
bracis-33560	279	5	(	(	PUNCT
bracis-33560	279	6	2020	2020	NUM
bracis-33560	279	7	)	)	PUNCT
bracis-33560	279	8	.	.	PUNCT
bracis-33560	280	1	https://doi.org/10.18653/v1/2020.acl-main.485	https://doi.org/10.18653/v1/2020.acl-main.485	NOUN
bracis-33560	280	2	bommasani	bommasani	NOUN
bracis-33560	280	3	,	,	PUNCT
bracis-33560	280	4	r.	r.	PROPN
bracis-33560	280	5	,	,	PUNCT
bracis-33560	280	6	et	et	PROPN
bracis-33560	280	7	 	 	SPACE
bracis-33560	280	8	al	al	PROPN
bracis-33560	280	9	.	.	PROPN
bracis-33560	280	10	:	:	PUNCT
bracis-33560	280	11	on	on	ADP
bracis-33560	280	12	the	the	DET
bracis-33560	280	13	opportunities	opportunity	NOUN
bracis-33560	280	14	and	and	CCONJ
bracis-33560	280	15	risks	risk	NOUN
bracis-33560	280	16	of	of	ADP
bracis-33560	280	17	foundation	foundation	NOUN
bracis-33560	280	18	models	model	NOUN
bracis-33560	280	19	.	.	PUNCT
bracis-33560	281	1	arxiv	arxiv	PROPN
bracis-33560	281	2	preprint	preprint	NOUN
bracis-33560	281	3	arxiv:2108.07258	arxiv:2108.07258	NUM
bracis-33560	281	4	(	(	PUNCT
bracis-33560	281	5	2021	2021	NUM
bracis-33560	281	6	)	)	PUNCT
bracis-33560	281	7	brown	brown	NOUN
bracis-33560	281	8	,	,	PUNCT
bracis-33560	281	9	t.	t.	PROPN
bracis-33560	281	10	,	,	PUNCT
bracis-33560	281	11	et	et	PROPN
bracis-33560	281	12	 	 	SPACE
bracis-33560	281	13	al	al	PROPN
bracis-33560	281	14	.	.	PROPN
bracis-33560	281	15	:	:	PUNCT
bracis-33560	282	1	language	language	NOUN
bracis-33560	282	2	models	model	NOUN
bracis-33560	282	3	are	be	AUX
bracis-33560	282	4	few	few	ADJ
bracis-33560	282	5	-	-	PUNCT
bracis-33560	282	6	shot	shot	NOUN
bracis-33560	282	7	learners	learner	NOUN
bracis-33560	282	8	.	.	PUNCT
bracis-33560	283	1	in	in	ADP
bracis-33560	283	2	:	:	PUNCT
bracis-33560	283	3	larochelle	larochelle	PROPN
bracis-33560	283	4	,	,	PUNCT
bracis-33560	283	5	h.	h.	PROPN
bracis-33560	283	6	,	,	PUNCT
bracis-33560	283	7	ranzato	ranzato	NOUN
bracis-33560	283	8	,	,	PUNCT
bracis-33560	283	9	m.	m.	NOUN
bracis-33560	283	10	,	,	PUNCT
bracis-33560	283	11	hadsell	hadsell	PROPN
bracis-33560	283	12	,	,	PUNCT
bracis-33560	283	13	r.	r.	PROPN
bracis-33560	283	14	,	,	PUNCT
bracis-33560	283	15	balcan	balcan	PROPN
bracis-33560	283	16	,	,	PUNCT
bracis-33560	283	17	m.	m.	NOUN
bracis-33560	283	18	,	,	PUNCT
bracis-33560	283	19	lin	lin	PROPN
bracis-33560	283	20	,	,	PUNCT
bracis-33560	283	21	h.	h.	PROPN
bracis-33560	283	22	(	(	PUNCT
bracis-33560	283	23	eds	eds	PROPN
bracis-33560	283	24	.	.	PUNCT
bracis-33560	283	25	)	)	PUNCT
bracis-33560	284	1	advances	advance	NOUN
bracis-33560	284	2	in	in	ADP
bracis-33560	284	3	neural	neural	ADJ
bracis-33560	284	4	information	information	NOUN
bracis-33560	284	5	processing	processing	NOUN
bracis-33560	284	6	systems	system	NOUN
bracis-33560	284	7	,	,	PUNCT
bracis-33560	284	8	vol	vol	NOUN
bracis-33560	284	9	.	.	PUNCT
bracis-33560	284	10	 	 	SPACE
bracis-33560	285	1	33	33	NUM
bracis-33560	285	2	,	,	PUNCT
bracis-33560	285	3	pp	pp	ADJ
bracis-33560	285	4	.	.	PUNCT
bracis-33560	285	5	1877–1901	1877–1901	NUM
bracis-33560	285	6	.	.	PUNCT
bracis-33560	286	1	curran	curran	PROPN
bracis-33560	286	2	associates	associates	PROPN
bracis-33560	286	3	,	,	PUNCT
bracis-33560	286	4	inc	inc	PROPN
bracis-33560	286	5	.	.	PROPN
bracis-33560	286	6	(	(	PUNCT
bracis-33560	286	7	2020	2020	NUM
bracis-33560	286	8	)	)	PUNCT
bracis-33560	286	9	.	.	PUNCT
bracis-33560	287	1	https://proceedings.neurips.cc/paper_files/paper/2020/file/1457c0d6bfcb4967418bfb8ac142f64a-paper.pdf	https://proceedings.neurips.cc/paper_files/paper/2020/file/1457c0d6bfcb4967418bfb8ac142f64a-paper.pdf	PROPN
bracis-33560	287	2	carmo	carmo	PROPN
bracis-33560	287	3	,	,	PUNCT
bracis-33560	287	4	d.	d.	PROPN
bracis-33560	287	5	,	,	PUNCT
bracis-33560	287	6	piau	piau	NOUN
bracis-33560	287	7	,	,	PUNCT
bracis-33560	287	8	m.	m.	NOUN
bracis-33560	287	9	,	,	PUNCT
bracis-33560	287	10	campiotti	campiotti	PROPN
bracis-33560	287	11	,	,	PUNCT
bracis-33560	287	12	i.	i.	PROPN
bracis-33560	287	13	,	,	PUNCT
bracis-33560	287	14	nogueira	nogueira	PROPN
bracis-33560	287	15	,	,	PUNCT
bracis-33560	287	16	r.	r.	PROPN
bracis-33560	287	17	,	,	PUNCT
bracis-33560	287	18	lotufo	lotufo	PROPN
bracis-33560	287	19	,	,	PUNCT
bracis-33560	287	20	r.	r.	PROPN
bracis-33560	287	21	:	:	PUNCT
bracis-33560	287	22	ptt5	ptt5	PROPN
bracis-33560	287	23	:	:	PUNCT
bracis-33560	287	24	pretraining	pretraine	VERB
bracis-33560	287	25	and	and	CCONJ
bracis-33560	287	26	validating	validate	VERB
bracis-33560	287	27	the	the	DET
bracis-33560	287	28	t5	t5	PROPN
bracis-33560	287	29	model	model	NOUN
bracis-33560	287	30	on	on	ADP
bracis-33560	287	31	brazilian	brazilian	ADJ
bracis-33560	287	32	portuguese	portuguese	ADJ
bracis-33560	287	33	data	datum	NOUN
bracis-33560	287	34	.	.	PUNCT
bracis-33560	288	1	arxiv	arxiv	PROPN
bracis-33560	288	2	preprint	preprint	PROPN
bracis-33560	288	3	arxiv:2008.09144	arxiv:2008.09144	ADP
bracis-33560	288	4	(	(	PUNCT
bracis-33560	288	5	2020	2020	NUM
bracis-33560	288	6	)	)	PUNCT
bracis-33560	288	7	computer	computer	NOUN
bracis-33560	288	8	,	,	PUNCT
bracis-33560	288	9	t.	t.	NOUN
bracis-33560	288	10	:	:	PUNCT
bracis-33560	288	11	redpajama	redpajama	NOUN
bracis-33560	288	12	:	:	PUNCT
bracis-33560	288	13	an	an	DET
bracis-33560	288	14	open	open	ADJ
bracis-33560	288	15	source	source	NOUN
bracis-33560	288	16	recipe	recipe	NOUN
bracis-33560	288	17	to	to	PART
bracis-33560	288	18	reproduce	reproduce	VERB
bracis-33560	288	19	llama	llama	NOUN
bracis-33560	288	20	training	training	NOUN
bracis-33560	288	21	dataset	dataset	NOUN
bracis-33560	288	22	(	(	PUNCT
bracis-33560	288	23	2023	2023	NUM
bracis-33560	288	24	)	)	PUNCT
bracis-33560	288	25	.	.	PUNCT
bracis-33560	289	1	https://github.com/togethercomputer/redpajama-data	https://github.com/togethercomputer/redpajama-data	PROPN
bracis-33560	289	2	crespo	crespo	PROPN
bracis-33560	289	3	,	,	PUNCT
bracis-33560	289	4	m.c.r.m	m.c.r.m	PROPN
bracis-33560	289	5	.	.	PROPN
bracis-33560	289	6	,	,	PUNCT
bracis-33560	289	7	et	et	PROPN
bracis-33560	289	8	 	 	SPACE
bracis-33560	289	9	al	al	PROPN
bracis-33560	289	10	.	.	PROPN
bracis-33560	289	11	:	:	PUNCT
bracis-33560	290	1	carolina	carolina	NOUN
bracis-33560	290	2	:	:	PUNCT
bracis-33560	290	3	a	a	DET
bracis-33560	290	4	general	general	ADJ
bracis-33560	290	5	corpus	corpus	NOUN
bracis-33560	290	6	of	of	ADP
bracis-33560	290	7	contemporary	contemporary	ADJ
bracis-33560	290	8	brazilian	brazilian	ADJ
bracis-33560	290	9	portuguese	portuguese	PROPN
bracis-33560	290	10	with	with	ADP
bracis-33560	290	11	provenance	provenance	NOUN
bracis-33560	290	12	,	,	PUNCT
bracis-33560	290	13	typology	typology	NOUN
bracis-33560	290	14	and	and	CCONJ
bracis-33560	290	15	versioning	versione	VERB
bracis-33560	290	16	information	information	NOUN
bracis-33560	290	17	,	,	PUNCT
bracis-33560	290	18	march	march	PROPN
bracis-33560	290	19	2023	2023	NUM
bracis-33560	290	20	google	google	PROPN
bracis-33560	290	21	scholar	scholar	NOUN
bracis-33560	290	22	  	  	SPACE
bracis-33560	290	23	finardi	finardi	PROPN
bracis-33560	290	24	,	,	PUNCT
bracis-33560	290	25	p.	p.	NOUN
bracis-33560	290	26	,	,	PUNCT
bracis-33560	290	27	viegas	viegas	PROPN
bracis-33560	290	28	,	,	PUNCT
bracis-33560	290	29	j.d	j.d	PROPN
bracis-33560	290	30	.	.	PROPN
bracis-33560	290	31	,	,	PUNCT
bracis-33560	290	32	ferreira	ferreira	PROPN
bracis-33560	290	33	,	,	PUNCT
bracis-33560	290	34	g.t	g.t	PROPN
bracis-33560	290	35	.	.	PROPN
bracis-33560	290	36	,	,	PUNCT
bracis-33560	290	37	mansano	mansano	PROPN
bracis-33560	290	38	,	,	PUNCT
bracis-33560	290	39	a.f	a.f	PROPN
bracis-33560	290	40	.	.	PROPN
bracis-33560	290	41	,	,	PUNCT
bracis-33560	290	42	caridá	caridá	PROPN
bracis-33560	290	43	,	,	PUNCT
bracis-33560	290	44	v.f	v.f	PROPN
bracis-33560	290	45	.	.	PUNCT
bracis-33560	290	46	:	:	PUNCT
bracis-33560	290	47	berta\(\backslash	berta\(\backslash	PROPN
bracis-33560	290	48	\)’u	\)’u	NUM
bracis-33560	290	49	:	:	PUNCT
bracis-33560	290	50	ita\(\backslash	ita\(\backslash	NUM
bracis-33560	290	51	\)’u	\)’u	NOUN
bracis-33560	290	52	bert	bert	NOUN
bracis-33560	290	53	for	for	ADP
bracis-33560	290	54	digital	digital	ADJ
bracis-33560	290	55	customer	customer	NOUN
bracis-33560	290	56	service	service	NOUN
bracis-33560	290	57	.	.	PUNCT
bracis-33560	291	1	arxiv	arxiv	PROPN
bracis-33560	291	2	preprint	preprint	PROPN
bracis-33560	291	3	arxiv:2101.12015	arxiv:2101.12015	NUM
bracis-33560	291	4	(	(	PUNCT
bracis-33560	291	5	2021	2021	NUM
bracis-33560	291	6	)	)	PUNCT
bracis-33560	291	7	gallegos	gallegos	PROPN
bracis-33560	291	8	,	,	PUNCT
bracis-33560	291	9	i.o	i.o	PROPN
bracis-33560	291	10	.	.	PROPN
bracis-33560	291	11	,	,	PUNCT
bracis-33560	291	12	et	et	PROPN
bracis-33560	291	13	 	 	SPACE
bracis-33560	291	14	al	al	PROPN
bracis-33560	291	15	.	.	PROPN
bracis-33560	291	16	:	:	PUNCT
bracis-33560	291	17	bias	bias	NOUN
bracis-33560	291	18	and	and	CCONJ
bracis-33560	291	19	fairness	fairness	NOUN
bracis-33560	291	20	in	in	ADP
bracis-33560	291	21	large	large	ADJ
bracis-33560	291	22	language	language	NOUN
bracis-33560	291	23	models	model	NOUN
bracis-33560	291	24	:	:	PUNCT
bracis-33560	291	25	a	a	DET
bracis-33560	291	26	survey	survey	NOUN
bracis-33560	291	27	.	.	PUNCT
bracis-33560	292	1	comput	comput	NOUN
bracis-33560	292	2	.	.	PUNCT
bracis-33560	293	1	linguist	linguist	NOUN
bracis-33560	293	2	.	.	PUNCT
bracis-33560	294	1	1–79	1–79	ADV
bracis-33560	294	2	(	(	PUNCT
bracis-33560	294	3	2024	2024	NUM
bracis-33560	294	4	)	)	PUNCT
bracis-33560	294	5	.	.	PUNCT
bracis-33560	295	1	https://doi.org/10.1162/coli_a_00524	https://doi.org/10.1162/coli_a_00524	PROPN
bracis-33560	295	2	gao	gao	PROPN
bracis-33560	295	3	,	,	PUNCT
bracis-33560	295	4	l.	l.	PROPN
bracis-33560	295	5	,	,	PUNCT
bracis-33560	295	6	et	et	PROPN
bracis-33560	295	7	 	 	SPACE
bracis-33560	295	8	al	al	PROPN
bracis-33560	295	9	.	.	PROPN
bracis-33560	295	10	:	:	PUNCT
bracis-33560	296	1	the	the	DET
bracis-33560	296	2	pile	pile	NOUN
bracis-33560	296	3	:	:	PUNCT
bracis-33560	296	4	an	an	DET
bracis-33560	296	5	800	800	NUM
bracis-33560	296	6	gb	gb	NOUN
bracis-33560	296	7	dataset	dataset	NOUN
bracis-33560	296	8	of	of	ADP
bracis-33560	296	9	diverse	diverse	ADJ
bracis-33560	296	10	text	text	NOUN
bracis-33560	296	11	for	for	ADP
bracis-33560	296	12	language	language	NOUN
bracis-33560	296	13	modeling	modeling	NOUN
bracis-33560	296	14	,	,	PUNCT
bracis-33560	296	15	december	december	PROPN
bracis-33560	296	16	2020	2020	NUM
bracis-33560	296	17	google	google	PROPN
bracis-33560	296	18	scholar	scholar	NOUN
bracis-33560	296	19	  	  	SPACE
bracis-33560	296	20	garcia	garcia	PROPN
bracis-33560	296	21	,	,	PUNCT
bracis-33560	296	22	g.l	g.l	PROPN
bracis-33560	296	23	.	.	PROPN
bracis-33560	296	24	,	,	PUNCT
bracis-33560	296	25	et	et	PROPN
bracis-33560	296	26	 	 	SPACE
bracis-33560	296	27	al	al	PROPN
bracis-33560	296	28	.	.	PROPN
bracis-33560	296	29	:	:	PUNCT
bracis-33560	296	30	introducing	introduce	VERB
bracis-33560	296	31	bode	bode	NOUN
bracis-33560	296	32	:	:	PUNCT
bracis-33560	296	33	a	a	DET
bracis-33560	296	34	fine	fine	ADV
bracis-33560	296	35	-	-	PUNCT
bracis-33560	296	36	tuned	tune	VERB
bracis-33560	296	37	large	large	ADJ
bracis-33560	296	38	language	language	NOUN
bracis-33560	296	39	model	model	NOUN
bracis-33560	296	40	for	for	ADP
bracis-33560	296	41	portuguese	portuguese	ADJ
bracis-33560	296	42	prompt	prompt	NOUN
bracis-33560	296	43	-	-	PUNCT
bracis-33560	296	44	based	base	VERB
bracis-33560	296	45	task	task	NOUN
bracis-33560	296	46	.	.	PUNCT
bracis-33560	297	1	arxiv	arxiv	PROPN
bracis-33560	297	2	preprint	preprint	VERB
bracis-33560	297	3	arxiv:2401.02909	arxiv:2401.02909	ADJ
bracis-33560	297	4	(	(	PUNCT
bracis-33560	297	5	2024	2024	NUM
bracis-33560	297	6	)	)	PUNCT
bracis-33560	297	7	garimella	garimella	PROPN
bracis-33560	297	8	,	,	PUNCT
bracis-33560	297	9	a.	a.	PROPN
bracis-33560	297	10	,	,	PUNCT
bracis-33560	297	11	mihalcea	mihalcea	PROPN
bracis-33560	297	12	,	,	PUNCT
bracis-33560	297	13	r.	r.	PROPN
bracis-33560	297	14	,	,	PUNCT
bracis-33560	297	15	amarnath	amarnath	PROPN
bracis-33560	297	16	,	,	PUNCT
bracis-33560	297	17	a.	a.	NOUN
bracis-33560	297	18	:	:	PUNCT
bracis-33560	297	19	demographic	demographic	ADJ
bracis-33560	297	20	-	-	PUNCT
bracis-33560	297	21	aware	aware	ADJ
bracis-33560	297	22	language	language	NOUN
bracis-33560	297	23	model	model	NOUN
bracis-33560	297	24	fine	fine	ADV
bracis-33560	297	25	-	-	PUNCT
bracis-33560	297	26	tuning	tuning	NOUN
bracis-33560	297	27	as	as	ADP
bracis-33560	297	28	a	a	DET
bracis-33560	297	29	bias	bias	NOUN
bracis-33560	297	30	mitigation	mitigation	NOUN
bracis-33560	297	31	technique	technique	NOUN
bracis-33560	297	32	.	.	PUNCT
bracis-33560	298	1	in	in	ADP
bracis-33560	298	2	:	:	PUNCT
bracis-33560	298	3	he	he	PRON
bracis-33560	298	4	,	,	PUNCT
bracis-33560	298	5	y.	y.	PROPN
bracis-33560	298	6	,	,	PUNCT
bracis-33560	298	7	et	et	PROPN
bracis-33560	298	8	 	 	SPACE
bracis-33560	298	9	al	al	PROPN
bracis-33560	298	10	.	.	PROPN
bracis-33560	298	11	(	(	PUNCT
bracis-33560	298	12	eds	ed	NOUN
bracis-33560	298	13	.	.	PUNCT
bracis-33560	298	14	)	)	PUNCT
bracis-33560	298	15	proceedings	proceeding	NOUN
bracis-33560	298	16	of	of	ADP
bracis-33560	298	17	the	the	DET
bracis-33560	298	18	2nd	2nd	ADJ
bracis-33560	298	19	conference	conference	NOUN
bracis-33560	298	20	of	of	ADP
bracis-33560	298	21	the	the	DET
bracis-33560	298	22	asia	asia	PROPN
bracis-33560	298	23	-	-	PUNCT
bracis-33560	298	24	pacific	pacific	PROPN
bracis-33560	298	25	chapter	chapter	NOUN
bracis-33560	298	26	of	of	ADP
bracis-33560	298	27	the	the	DET
bracis-33560	298	28	association	association	NOUN
bracis-33560	298	29	for	for	ADP
bracis-33560	298	30	computational	computational	ADJ
bracis-33560	298	31	linguistics	linguistic	NOUN
bracis-33560	298	32	and	and	CCONJ
bracis-33560	298	33	the	the	DET
bracis-33560	298	34	12th	12th	ADJ
bracis-33560	298	35	international	international	ADJ
bracis-33560	298	36	joint	joint	ADJ
bracis-33560	298	37	conference	conference	NOUN
bracis-33560	298	38	on	on	ADP
bracis-33560	298	39	natural	natural	ADJ
bracis-33560	298	40	language	language	NOUN
bracis-33560	298	41	processing	processing	NOUN
bracis-33560	298	42	(	(	PUNCT
bracis-33560	298	43	volume	volume	NOUN
bracis-33560	298	44	2	2	NUM
bracis-33560	298	45	:	:	PUNCT
bracis-33560	298	46	short	short	ADJ
bracis-33560	298	47	papers	paper	NOUN
bracis-33560	298	48	)	)	PUNCT
bracis-33560	298	49	,	,	PUNCT
bracis-33560	298	50	pp	pp	ADP
bracis-33560	298	51	.	.	PUNCT
bracis-33560	299	1	311–319	311–319	NUM
bracis-33560	299	2	.	.	PUNCT
bracis-33560	300	1	association	association	NOUN
bracis-33560	300	2	for	for	ADP
bracis-33560	300	3	computational	computational	ADJ
bracis-33560	300	4	linguistics	linguistic	NOUN
bracis-33560	300	5	,	,	PUNCT
bracis-33560	300	6	online	online	ADV
bracis-33560	300	7	only	only	ADV
bracis-33560	300	8	,	,	PUNCT
bracis-33560	300	9	november	november	PROPN
bracis-33560	300	10	2022	2022	NUM
bracis-33560	300	11	.	.	PUNCT
bracis-33560	301	1	https://aclanthology.org/2022.aacl-short.38	https://aclanthology.org/2022.aacl-short.38	PROPN
bracis-33560	301	2	habernal	habernal	PROPN
bracis-33560	301	3	,	,	PUNCT
bracis-33560	301	4	i.	i.	PROPN
bracis-33560	301	5	,	,	PUNCT
bracis-33560	301	6	zayed	zayed	PROPN
bracis-33560	301	7	,	,	PUNCT
bracis-33560	301	8	o.	o.	NOUN
bracis-33560	301	9	,	,	PUNCT
bracis-33560	301	10	gurevych	gurevych	NOUN
bracis-33560	301	11	,	,	PUNCT
bracis-33560	301	12	i.	i.	NOUN
bracis-33560	301	13	:	:	PUNCT
bracis-33560	301	14	c4corpus	c4corpus	PROPN
bracis-33560	301	15	:	:	PUNCT
bracis-33560	301	16	multilingual	multilingual	ADJ
bracis-33560	301	17	web	web	NOUN
bracis-33560	301	18	-	-	NOUN
bracis-33560	301	19	size	size	NOUN
bracis-33560	301	20	corpus	corpus	NOUN
bracis-33560	301	21	with	with	ADP
bracis-33560	301	22	free	free	ADJ
bracis-33560	301	23	license	license	NOUN
bracis-33560	301	24	.	.	PUNCT
bracis-33560	302	1	in	in	ADP
bracis-33560	302	2	:	:	PUNCT
bracis-33560	302	3	calzolari	calzolari	PROPN
bracis-33560	302	4	,	,	PUNCT
bracis-33560	302	5	n.	n.	PROPN
bracis-33560	302	6	,	,	PUNCT
bracis-33560	302	7	et	et	PROPN
bracis-33560	302	8	 	 	SPACE
bracis-33560	302	9	al	al	PROPN
bracis-33560	302	10	.	.	PROPN
bracis-33560	302	11	(	(	PUNCT
bracis-33560	302	12	eds	ed	NOUN
bracis-33560	302	13	.	.	PUNCT
bracis-33560	302	14	)	)	PUNCT
bracis-33560	302	15	proceedings	proceeding	NOUN
bracis-33560	302	16	of	of	ADP
bracis-33560	302	17	the	the	DET
bracis-33560	302	18	tenth	tenth	ADJ
bracis-33560	302	19	international	international	ADJ
bracis-33560	302	20	conference	conference	NOUN
bracis-33560	302	21	on	on	ADP
bracis-33560	302	22	language	language	NOUN
bracis-33560	302	23	resources	resource	NOUN
bracis-33560	302	24	and	and	CCONJ
bracis-33560	302	25	evaluation	evaluation	NOUN
bracis-33560	302	26	(	(	PUNCT
bracis-33560	302	27	lrec	lrec	PROPN
bracis-33560	302	28	2016	2016	NUM
bracis-33560	302	29	)	)	PUNCT
bracis-33560	302	30	,	,	PUNCT
bracis-33560	302	31	pp	pp	ADP
bracis-33560	302	32	.	.	PUNCT
bracis-33560	303	1	914–922	914–922	NUM
bracis-33560	303	2	.	.	PUNCT
bracis-33560	303	3	european	european	ADJ
bracis-33560	303	4	language	language	PROPN
bracis-33560	303	5	resources	resource	NOUN
bracis-33560	303	6	association	association	NOUN
bracis-33560	303	7	(	(	PUNCT
bracis-33560	303	8	elra	elra	ADV
bracis-33560	303	9	)	)	PUNCT
bracis-33560	303	10	,	,	PUNCT
bracis-33560	303	11	portorož	portorož	NOUN
bracis-33560	303	12	,	,	PUNCT
bracis-33560	303	13	slovenia	slovenia	PROPN
bracis-33560	303	14	,	,	PUNCT
bracis-33560	303	15	may	may	PROPN
bracis-33560	303	16	2016	2016	NUM
bracis-33560	303	17	.	.	PUNCT
bracis-33560	304	1	https://aclanthology.org/l16-1146	https://aclanthology.org/l16-1146	PROPN
bracis-33560	304	2	hoffmann	hoffmann	PROPN
bracis-33560	304	3	,	,	PUNCT
bracis-33560	304	4	j.	j.	PROPN
bracis-33560	304	5	,	,	PUNCT
bracis-33560	304	6	et	et	PROPN
bracis-33560	304	7	 	 	SPACE
bracis-33560	304	8	al	al	PROPN
bracis-33560	304	9	.	.	PROPN
bracis-33560	304	10	:	:	PUNCT
bracis-33560	305	1	an	an	DET
bracis-33560	305	2	empirical	empirical	ADJ
bracis-33560	305	3	analysis	analysis	NOUN
bracis-33560	305	4	of	of	ADP
bracis-33560	305	5	compute	compute	NOUN
bracis-33560	305	6	-	-	PUNCT
bracis-33560	305	7	optimal	optimal	ADJ
bracis-33560	305	8	large	large	ADJ
bracis-33560	305	9	language	language	NOUN
bracis-33560	305	10	model	model	NOUN
bracis-33560	305	11	training	training	NOUN
bracis-33560	305	12	.	.	PUNCT
bracis-33560	306	1	in	in	ADP
bracis-33560	306	2	:	:	PUNCT
bracis-33560	306	3	koyejo	koyejo	NOUN
bracis-33560	306	4	,	,	PUNCT
bracis-33560	306	5	s.	s.	PROPN
bracis-33560	306	6	,	,	PUNCT
bracis-33560	306	7	mohamed	mohamed	PROPN
bracis-33560	306	8	,	,	PUNCT
bracis-33560	306	9	s.	s.	PROPN
bracis-33560	306	10	,	,	PUNCT
bracis-33560	306	11	agarwal	agarwal	PROPN
bracis-33560	306	12	,	,	PUNCT
bracis-33560	306	13	a.	a.	PROPN
bracis-33560	306	14	,	,	PUNCT
bracis-33560	306	15	belgrave	belgrave	PROPN
bracis-33560	306	16	,	,	PUNCT
bracis-33560	306	17	d.	d.	PROPN
bracis-33560	306	18	,	,	PUNCT
bracis-33560	306	19	cho	cho	PROPN
bracis-33560	306	20	,	,	PUNCT
bracis-33560	306	21	k.	k.	PROPN
bracis-33560	306	22	,	,	PUNCT
bracis-33560	306	23	oh	oh	INTJ
bracis-33560	306	24	,	,	PUNCT
bracis-33560	306	25	a.	a.	NOUN
bracis-33560	306	26	(	(	PUNCT
bracis-33560	306	27	eds	eds	PROPN
bracis-33560	306	28	.	.	PUNCT
bracis-33560	306	29	)	)	PUNCT
bracis-33560	306	30	advances	advance	NOUN
bracis-33560	306	31	in	in	ADP
bracis-33560	306	32	neural	neural	ADJ
bracis-33560	306	33	information	information	NOUN
bracis-33560	306	34	processing	processing	NOUN
bracis-33560	306	35	systems	system	NOUN
bracis-33560	306	36	,	,	PUNCT
bracis-33560	306	37	vol	vol	NOUN
bracis-33560	306	38	.	.	PUNCT
bracis-33560	306	39	 	 	SPACE
bracis-33560	306	40	35	35	NUM
bracis-33560	306	41	,	,	PUNCT
bracis-33560	306	42	pp	pp	ADJ
bracis-33560	306	43	.	.	PUNCT
bracis-33560	307	1	30016–30030	30016–30030	NUM
bracis-33560	307	2	.	.	PUNCT
bracis-33560	308	1	curran	curran	PROPN
bracis-33560	308	2	associates	associates	PROPN
bracis-33560	308	3	,	,	PUNCT
bracis-33560	308	4	inc	inc	PROPN
bracis-33560	308	5	.	.	PROPN
bracis-33560	308	6	(	(	PUNCT
bracis-33560	308	7	2022	2022	NUM
bracis-33560	308	8	)	)	PUNCT
bracis-33560	308	9	.	.	PUNCT
bracis-33560	309	1	https://proceedings.neurips.cc/paper_files/paper/2022/file/c1e2faff6f588870935f114ebe04a3e5-paper-conference.pdf	https://proceedings.neurips.cc/paper_files/paper/2022/file/c1e2faff6f588870935f114ebe04a3e5-paper-conference.pdf	PROPN
bracis-33560	309	2	jiang	jiang	PROPN
bracis-33560	309	3	,	,	PUNCT
bracis-33560	309	4	a.q	a.q	PROPN
bracis-33560	309	5	.	.	PROPN
bracis-33560	309	6	,	,	PUNCT
bracis-33560	309	7	et	et	PROPN
bracis-33560	309	8	 	 	SPACE
bracis-33560	309	9	al	al	PROPN
bracis-33560	309	10	.	.	PUNCT
bracis-33560	309	11	:	:	PUNCT
bracis-33560	310	1	mistral	mistral	ADJ
bracis-33560	310	2	7b	7b	PROPN
bracis-33560	310	3	.	.	PUNCT
bracis-33560	311	1	arxiv	arxiv	PROPN
bracis-33560	311	2	preprint	preprint	VERB
bracis-33560	311	3	arxiv:2310.06825	arxiv:2310.06825	NOUN
bracis-33560	311	4	(	(	PUNCT
bracis-33560	311	5	2023	2023	NUM
bracis-33560	311	6	)	)	PUNCT
bracis-33560	312	1	joulin	joulin	PROPN
bracis-33560	312	2	,	,	PUNCT
bracis-33560	312	3	a.	a.	NOUN
bracis-33560	312	4	,	,	PUNCT
bracis-33560	312	5	grave	grave	NOUN
bracis-33560	312	6	,	,	PUNCT
bracis-33560	312	7	e.	e.	PROPN
bracis-33560	312	8	,	,	PUNCT
bracis-33560	312	9	bojanowski	bojanowski	VERB
bracis-33560	312	10	,	,	PUNCT
bracis-33560	312	11	p.	p.	NOUN
bracis-33560	312	12	,	,	PUNCT
bracis-33560	312	13	mikolov	mikolov	PROPN
bracis-33560	312	14	,	,	PUNCT
bracis-33560	312	15	t.	t.	NOUN
bracis-33560	312	16	:	:	PUNCT
bracis-33560	312	17	bag	bag	NOUN
bracis-33560	312	18	of	of	ADP
bracis-33560	312	19	tricks	trick	NOUN
bracis-33560	312	20	for	for	ADP
bracis-33560	312	21	efficient	efficient	ADJ
bracis-33560	312	22	text	text	NOUN
bracis-33560	312	23	classification	classification	NOUN
bracis-33560	312	24	.	.	PUNCT
bracis-33560	313	1	in	in	ADP
bracis-33560	313	2	:	:	PUNCT
bracis-33560	313	3	proceedings	proceeding	NOUN
bracis-33560	313	4	of	of	ADP
bracis-33560	313	5	the	the	DET
bracis-33560	313	6	15th	15th	ADJ
bracis-33560	313	7	conference	conference	NOUN
bracis-33560	313	8	of	of	ADP
bracis-33560	313	9	the	the	DET
bracis-33560	313	10	european	european	ADJ
bracis-33560	313	11	chapter	chapter	NOUN
bracis-33560	313	12	of	of	ADP
bracis-33560	313	13	the	the	DET
bracis-33560	313	14	association	association	NOUN
bracis-33560	313	15	for	for	ADP
bracis-33560	313	16	computational	computational	ADJ
bracis-33560	313	17	linguistics	linguistic	NOUN
bracis-33560	313	18	:	:	PUNCT
bracis-33560	313	19	volume	volume	NOUN
bracis-33560	313	20	2	2	NUM
bracis-33560	313	21	,	,	PUNCT
bracis-33560	313	22	short	short	ADJ
bracis-33560	313	23	papers	paper	NOUN
bracis-33560	313	24	,	,	PUNCT
bracis-33560	313	25	pp	pp	ADP
bracis-33560	313	26	.	.	PUNCT
bracis-33560	314	1	427–431	427–431	NUM
bracis-33560	314	2	.	.	PUNCT
bracis-33560	315	1	association	association	NOUN
bracis-33560	315	2	for	for	ADP
bracis-33560	315	3	computational	computational	ADJ
bracis-33560	315	4	linguistics	linguistic	NOUN
bracis-33560	315	5	,	,	PUNCT
bracis-33560	315	6	april	april	PROPN
bracis-33560	315	7	2017	2017	NUM
bracis-33560	315	8	google	google	PROPN
bracis-33560	315	9	scholar	scholar	NOUN
bracis-33560	315	10	  	  	SPACE
bracis-33560	315	11	kenton	kenton	PROPN
bracis-33560	315	12	,	,	PUNCT
bracis-33560	315	13	j.d.m.w.c	j.d.m.w.c	NOUN
bracis-33560	315	14	.	.	PROPN
bracis-33560	315	15	,	,	PUNCT
bracis-33560	315	16	toutanova	toutanova	PROPN
bracis-33560	315	17	,	,	PUNCT
bracis-33560	315	18	l.k	l.k	PROPN
bracis-33560	315	19	.	.	PROPN
bracis-33560	315	20	:	:	PUNCT
bracis-33560	316	1	bert	bert	PROPN
bracis-33560	316	2	:	:	PUNCT
bracis-33560	316	3	pre	pre	ADJ
bracis-33560	316	4	-	-	NOUN
bracis-33560	316	5	training	training	NOUN
bracis-33560	316	6	of	of	ADP
bracis-33560	316	7	deep	deep	ADJ
bracis-33560	316	8	bidirectional	bidirectional	ADJ
bracis-33560	316	9	transformers	transformer	NOUN
bracis-33560	316	10	for	for	ADP
bracis-33560	316	11	language	language	NOUN
bracis-33560	316	12	understanding	understanding	NOUN
bracis-33560	316	13	.	.	PUNCT
bracis-33560	317	1	in	in	ADP
bracis-33560	317	2	:	:	PUNCT
bracis-33560	317	3	proceedings	proceeding	NOUN
bracis-33560	317	4	of	of	ADP
bracis-33560	317	5	naacl	naacl	PROPN
bracis-33560	317	6	-	-	PUNCT
bracis-33560	317	7	hlt	hlt	PROPN
bracis-33560	317	8	,	,	PUNCT
bracis-33560	317	9	vol	vol	NOUN
bracis-33560	317	10	.	.	PUNCT
bracis-33560	317	11	 	 	SPACE
bracis-33560	317	12	1	1	NUM
bracis-33560	317	13	,	,	PUNCT
bracis-33560	317	14	p.	p.	NOUN
bracis-33560	317	15	 	 	SPACE
bracis-33560	317	16	2	2	NUM
bracis-33560	317	17	(	(	PUNCT
bracis-33560	317	18	2019	2019	NUM
bracis-33560	317	19	)	)	PUNCT
bracis-33560	317	20	google	google	NOUN
bracis-33560	317	21	scholar	scholar	NOUN
bracis-33560	317	22	  	  	SPACE
bracis-33560	317	23	larcher	larcher	ADV
bracis-33560	317	24	,	,	PUNCT
bracis-33560	317	25	c.	c.	PROPN
bracis-33560	317	26	,	,	PUNCT
bracis-33560	317	27	piau	piau	NOUN
bracis-33560	317	28	,	,	PUNCT
bracis-33560	317	29	m.	m.	NOUN
bracis-33560	317	30	,	,	PUNCT
bracis-33560	317	31	finardi	finardi	PROPN
bracis-33560	317	32	,	,	PUNCT
bracis-33560	317	33	p.	p.	NOUN
bracis-33560	317	34	,	,	PUNCT
bracis-33560	317	35	gengo	gengo	PROPN
bracis-33560	317	36	,	,	PUNCT
bracis-33560	317	37	p.	p.	PROPN
bracis-33560	317	38	,	,	PUNCT
bracis-33560	317	39	esposito	esposito	PROPN
bracis-33560	317	40	,	,	PUNCT
bracis-33560	317	41	p.	p.	PROPN
bracis-33560	317	42	,	,	PUNCT
bracis-33560	317	43	caridá	caridá	NOUN
bracis-33560	317	44	,	,	PUNCT
bracis-33560	317	45	v.	v.	ADV
bracis-33560	317	46	:	:	PUNCT
bracis-33560	317	47	cabrita	cabrita	NOUN
bracis-33560	317	48	:	:	PUNCT
bracis-33560	317	49	closing	close	VERB
bracis-33560	317	50	the	the	DET
bracis-33560	317	51	gap	gap	NOUN
bracis-33560	317	52	for	for	ADP
bracis-33560	317	53	foreign	foreign	ADJ
bracis-33560	317	54	languages	language	NOUN
bracis-33560	317	55	.	.	PUNCT
bracis-33560	318	1	arxiv	arxiv	PROPN
bracis-33560	318	2	preprint	preprint	VERB
bracis-33560	318	3	arxiv:2308.11878	arxiv:2308.11878	ADJ
bracis-33560	318	4	(	(	PUNCT
bracis-33560	318	5	2023	2023	NUM
bracis-33560	318	6	)	)	PUNCT
bracis-33560	318	7	liu	liu	PROPN
bracis-33560	318	8	,	,	PUNCT
bracis-33560	318	9	y.	y.	PROPN
bracis-33560	318	10	,	,	PUNCT
bracis-33560	318	11	cao	cao	PROPN
bracis-33560	318	12	,	,	PUNCT
bracis-33560	318	13	j.	j.	PROPN
bracis-33560	318	14	,	,	PUNCT
bracis-33560	318	15	liu	liu	PROPN
bracis-33560	318	16	,	,	PUNCT
bracis-33560	318	17	c.	c.	PROPN
bracis-33560	318	18	,	,	PUNCT
bracis-33560	318	19	ding	ding	NOUN
bracis-33560	318	20	,	,	PUNCT
bracis-33560	318	21	k.	k.	PROPN
bracis-33560	318	22	,	,	PUNCT
bracis-33560	318	23	jin	jin	PROPN
bracis-33560	318	24	,	,	PUNCT
bracis-33560	318	25	l.	l.	PROPN
bracis-33560	318	26	:	:	PUNCT
bracis-33560	318	27	datasets	dataset	NOUN
bracis-33560	318	28	for	for	ADP
bracis-33560	318	29	large	large	ADJ
bracis-33560	318	30	language	language	NOUN
bracis-33560	318	31	models	model	NOUN
bracis-33560	318	32	:	:	PUNCT
bracis-33560	318	33	a	a	DET
bracis-33560	318	34	comprehensive	comprehensive	ADJ
bracis-33560	318	35	survey	survey	NOUN
bracis-33560	318	36	.	.	PUNCT
bracis-33560	319	1	arxiv	arxiv	PROPN
bracis-33560	319	2	preprint	preprint	NOUN
bracis-33560	319	3	arxiv:2402.18041	arxiv:2402.18041	NOUN
bracis-33560	319	4	(	(	PUNCT
bracis-33560	319	5	2024	2024	NUM
bracis-33560	319	6	)	)	PUNCT
bracis-33560	319	7	mei	mei	PROPN
bracis-33560	319	8	,	,	PUNCT
bracis-33560	319	9	k.	k.	PROPN
bracis-33560	319	10	,	,	PUNCT
bracis-33560	319	11	fereidooni	fereidooni	PROPN
bracis-33560	319	12	,	,	PUNCT
bracis-33560	319	13	s.	s.	PROPN
bracis-33560	319	14	,	,	PUNCT
bracis-33560	319	15	caliskan	caliskan	PROPN
bracis-33560	319	16	,	,	PUNCT
bracis-33560	319	17	a.	a.	NOUN
bracis-33560	319	18	:	:	PUNCT
bracis-33560	319	19	bias	bias	NOUN
bracis-33560	319	20	against	against	ADP
bracis-33560	319	21	93	93	NUM
bracis-33560	319	22	stigmatized	stigmatize	VERB
bracis-33560	319	23	groups	group	NOUN
bracis-33560	319	24	in	in	ADP
bracis-33560	319	25	masked	masked	ADJ
bracis-33560	319	26	language	language	NOUN
bracis-33560	319	27	models	model	NOUN
bracis-33560	319	28	and	and	CCONJ
bracis-33560	319	29	downstream	downstream	ADJ
bracis-33560	319	30	sentiment	sentiment	NOUN
bracis-33560	319	31	classification	classification	NOUN
bracis-33560	319	32	tasks	task	NOUN
bracis-33560	319	33	.	.	PUNCT
bracis-33560	320	1	in	in	ADP
bracis-33560	320	2	:	:	PUNCT
bracis-33560	320	3	2023	2023	NUM
bracis-33560	320	4	acm	acm	NOUN
bracis-33560	320	5	conference	conference	NOUN
bracis-33560	320	6	on	on	ADP
bracis-33560	320	7	fairness	fairness	NOUN
bracis-33560	320	8	,	,	PUNCT
bracis-33560	320	9	accountability	accountability	NOUN
bracis-33560	320	10	,	,	PUNCT
bracis-33560	320	11	and	and	CCONJ
bracis-33560	320	12	transparency	transparency	NOUN
bracis-33560	320	13	,	,	PUNCT
bracis-33560	320	14	facct	facct	ADJ
bracis-33560	320	15	2023	2023	NUM
bracis-33560	320	16	.	.	PUNCT
bracis-33560	320	17	acm	acm	PROPN
bracis-33560	320	18	,	,	PUNCT
bracis-33560	320	19	june	june	PROPN
bracis-33560	320	20	2023	2023	NUM
bracis-33560	320	21	.	.	PUNCT
bracis-33560	321	1	https://doi.org/10.1145/3593013.3594109	https://doi.org/10.1145/3593013.3594109	PROPN
bracis-33560	321	2	nadeem	nadeem	PROPN
bracis-33560	321	3	,	,	PUNCT
bracis-33560	321	4	m.	m.	NOUN
bracis-33560	321	5	,	,	PUNCT
bracis-33560	321	6	bethke	bethke	PROPN
bracis-33560	321	7	,	,	PUNCT
bracis-33560	321	8	a.	a.	NOUN
bracis-33560	321	9	,	,	PUNCT
bracis-33560	321	10	reddy	reddy	PROPN
bracis-33560	321	11	,	,	PUNCT
bracis-33560	321	12	s.	s.	PROPN
bracis-33560	321	13	:	:	PUNCT
bracis-33560	321	14	stereoset	stereoset	PROPN
bracis-33560	321	15	:	:	PUNCT
bracis-33560	321	16	measuring	measure	VERB
bracis-33560	321	17	stereotypical	stereotypical	ADJ
bracis-33560	321	18	bias	bias	NOUN
bracis-33560	321	19	in	in	ADP
bracis-33560	321	20	pretrained	pretraine	VERB
bracis-33560	321	21	language	language	NOUN
bracis-33560	321	22	models	model	NOUN
bracis-33560	321	23	.	.	PUNCT
bracis-33560	322	1	in	in	ADP
bracis-33560	322	2	:	:	PUNCT
bracis-33560	322	3	proceedings	proceeding	NOUN
bracis-33560	322	4	of	of	ADP
bracis-33560	322	5	the	the	DET
bracis-33560	322	6	59th	59th	ADJ
bracis-33560	322	7	annual	annual	ADJ
bracis-33560	322	8	meeting	meeting	NOUN
bracis-33560	322	9	of	of	ADP
bracis-33560	322	10	the	the	DET
bracis-33560	322	11	association	association	NOUN
bracis-33560	322	12	for	for	ADP
bracis-33560	322	13	computational	computational	ADJ
bracis-33560	322	14	linguistics	linguistic	NOUN
bracis-33560	322	15	and	and	CCONJ
bracis-33560	322	16	the	the	DET
bracis-33560	322	17	11th	11th	ADJ
bracis-33560	322	18	international	international	ADJ
bracis-33560	322	19	joint	joint	ADJ
bracis-33560	322	20	conference	conference	NOUN
bracis-33560	322	21	on	on	ADP
bracis-33560	322	22	natural	natural	ADJ
bracis-33560	322	23	language	language	NOUN
bracis-33560	322	24	processing	processing	NOUN
bracis-33560	322	25	(	(	PUNCT
bracis-33560	322	26	volume	volume	NOUN
bracis-33560	322	27	1	1	NUM
bracis-33560	322	28	:	:	PUNCT
bracis-33560	322	29	long	long	ADJ
bracis-33560	322	30	papers	paper	NOUN
bracis-33560	322	31	)	)	PUNCT
bracis-33560	322	32	.	.	PUNCT
bracis-33560	323	1	association	association	NOUN
bracis-33560	323	2	for	for	ADP
bracis-33560	323	3	computational	computational	ADJ
bracis-33560	323	4	linguistics	linguistic	NOUN
bracis-33560	323	5	(	(	PUNCT
bracis-33560	323	6	2021	2021	NUM
bracis-33560	323	7	)	)	PUNCT
bracis-33560	323	8	.	.	PUNCT
bracis-33560	324	1	https://doi.org/10.18653/v1/2021.acl-long.416	https://doi.org/10.18653/v1/2021.acl-long.416	PROPN
bracis-33560	324	2	nangia	nangia	PROPN
bracis-33560	324	3	,	,	PUNCT
bracis-33560	324	4	n.	n.	NOUN
bracis-33560	324	5	,	,	PUNCT
bracis-33560	324	6	vania	vania	PROPN
bracis-33560	324	7	,	,	PUNCT
bracis-33560	324	8	c.	c.	PROPN
bracis-33560	324	9	,	,	PUNCT
bracis-33560	324	10	bhalerao	bhalerao	PROPN
bracis-33560	324	11	,	,	PUNCT
bracis-33560	324	12	r.	r.	PROPN
bracis-33560	324	13	,	,	PUNCT
bracis-33560	324	14	bowman	bowman	PROPN
bracis-33560	324	15	,	,	PUNCT
bracis-33560	324	16	s.r	s.r	PROPN
bracis-33560	324	17	.	.	PROPN
bracis-33560	324	18	:	:	PUNCT
bracis-33560	325	1	crows	crow	NOUN
bracis-33560	325	2	-	-	PUNCT
bracis-33560	325	3	pairs	pair	NOUN
bracis-33560	325	4	:	:	PUNCT
bracis-33560	325	5	a	a	DET
bracis-33560	325	6	challenge	challenge	NOUN
bracis-33560	325	7	dataset	dataset	VERB
bracis-33560	325	8	for	for	ADP
bracis-33560	325	9	measuring	measure	VERB
bracis-33560	325	10	social	social	ADJ
bracis-33560	325	11	biases	bias	NOUN
bracis-33560	325	12	in	in	ADP
bracis-33560	325	13	masked	masked	ADJ
bracis-33560	325	14	language	language	NOUN
bracis-33560	325	15	models	model	NOUN
bracis-33560	325	16	.	.	PUNCT
bracis-33560	326	1	in	in	ADP
bracis-33560	326	2	:	:	PUNCT
bracis-33560	326	3	proceedings	proceeding	NOUN
bracis-33560	326	4	of	of	ADP
bracis-33560	326	5	the	the	DET
bracis-33560	326	6	2020	2020	NUM
bracis-33560	326	7	conference	conference	NOUN
bracis-33560	326	8	on	on	ADP
bracis-33560	326	9	empirical	empirical	ADJ
bracis-33560	326	10	methods	method	NOUN
bracis-33560	326	11	in	in	ADP
bracis-33560	326	12	natural	natural	ADJ
bracis-33560	326	13	language	language	NOUN
bracis-33560	326	14	processing	processing	NOUN
bracis-33560	326	15	(	(	PUNCT
bracis-33560	326	16	emnlp	emnlp	ADJ
bracis-33560	326	17	)	)	PUNCT
bracis-33560	326	18	.	.	PUNCT
bracis-33560	327	1	association	association	NOUN
bracis-33560	327	2	for	for	ADP
bracis-33560	327	3	computational	computational	ADJ
bracis-33560	327	4	linguistics	linguistic	NOUN
bracis-33560	327	5	(	(	PUNCT
bracis-33560	327	6	2020	2020	NUM
bracis-33560	327	7	)	)	PUNCT
bracis-33560	327	8	.	.	PUNCT
bracis-33560	328	1	https://doi.org/10.18653/v1/2020.emnlp-main.154	https://doi.org/10.18653/v1/2020.emnlp-main.154	VERB
bracis-33560	328	2	overwijk	overwijk	NOUN
bracis-33560	328	3	,	,	PUNCT
bracis-33560	328	4	a.	a.	NOUN
bracis-33560	328	5	,	,	PUNCT
bracis-33560	328	6	xiong	xiong	PROPN
bracis-33560	328	7	,	,	PUNCT
bracis-33560	328	8	c.	c.	PROPN
bracis-33560	328	9	,	,	PUNCT
bracis-33560	328	10	callan	callan	PROPN
bracis-33560	328	11	,	,	PUNCT
bracis-33560	328	12	j.	j.	PROPN
bracis-33560	328	13	:	:	PUNCT
bracis-33560	328	14	clueweb22	clueweb22	PROPN
bracis-33560	328	15	:	:	PUNCT
bracis-33560	328	16	10	10	NUM
bracis-33560	328	17	billion	billion	NUM
bracis-33560	328	18	web	web	NOUN
bracis-33560	328	19	documents	document	NOUN
bracis-33560	328	20	with	with	ADP
bracis-33560	328	21	rich	rich	ADJ
bracis-33560	328	22	information	information	NOUN
bracis-33560	328	23	.	.	PUNCT
bracis-33560	329	1	in	in	ADP
bracis-33560	329	2	:	:	PUNCT
bracis-33560	329	3	proceedings	proceeding	NOUN
bracis-33560	329	4	of	of	ADP
bracis-33560	329	5	the	the	DET
bracis-33560	329	6	45th	45th	ADJ
bracis-33560	329	7	international	international	PROPN
bracis-33560	329	8	acm	acm	PROPN
bracis-33560	329	9	sigir	sigir	PROPN
bracis-33560	329	10	conference	conference	PROPN
bracis-33560	329	11	on	on	ADP
bracis-33560	329	12	research	research	NOUN
bracis-33560	329	13	and	and	CCONJ
bracis-33560	329	14	development	development	NOUN
bracis-33560	329	15	in	in	ADP
bracis-33560	329	16	information	information	NOUN
bracis-33560	329	17	retrieval	retrieval	NOUN
bracis-33560	329	18	,	,	PUNCT
bracis-33560	329	19	pp	pp	ADP
bracis-33560	329	20	.	.	PUNCT
bracis-33560	330	1	3360–3362	3360–3362	NUM
bracis-33560	330	2	(	(	PUNCT
bracis-33560	330	3	2022	2022	NUM
bracis-33560	330	4	)	)	PUNCT
bracis-33560	330	5	google	google	PROPN
bracis-33560	330	6	scholar	scholar	NOUN
bracis-33560	330	7	  	  	SPACE
bracis-33560	330	8	pires	pire	NOUN
bracis-33560	330	9	,	,	PUNCT
bracis-33560	330	10	r.	r.	PROPN
bracis-33560	330	11	,	,	PUNCT
bracis-33560	330	12	abonizio	abonizio	PROPN
bracis-33560	330	13	,	,	PUNCT
bracis-33560	330	14	h.	h.	PROPN
bracis-33560	330	15	,	,	PUNCT
bracis-33560	330	16	almeida	almeida	PROPN
bracis-33560	330	17	,	,	PUNCT
bracis-33560	330	18	t.s	t.s	PROPN
bracis-33560	330	19	.	.	PROPN
bracis-33560	330	20	,	,	PUNCT
bracis-33560	330	21	nogueira	nogueira	PROPN
bracis-33560	330	22	,	,	PUNCT
bracis-33560	330	23	r.	r.	PROPN
bracis-33560	330	24	:	:	PUNCT
bracis-33560	330	25	sabiá	sabiá	NOUN
bracis-33560	330	26	:	:	PUNCT
bracis-33560	330	27	portuguese	portuguese	VERB
bracis-33560	330	28	large	large	ADJ
bracis-33560	330	29	language	language	NOUN
bracis-33560	330	30	models	model	NOUN
bracis-33560	330	31	.	.	PUNCT
bracis-33560	331	1	in	in	ADP
bracis-33560	331	2	:	:	PUNCT
bracis-33560	331	3	brazilian	brazilian	ADJ
bracis-33560	331	4	conference	conference	NOUN
bracis-33560	331	5	on	on	ADP
bracis-33560	331	6	intelligent	intelligent	ADJ
bracis-33560	331	7	systems	system	NOUN
bracis-33560	331	8	,	,	PUNCT
bracis-33560	331	9	pp	pp	ADV
bracis-33560	331	10	.	.	PUNCT
bracis-33560	332	1	226–240	226–240	NUM
bracis-33560	332	2	(	(	PUNCT
bracis-33560	332	3	2023	2023	NUM
bracis-33560	332	4	)	)	PUNCT
bracis-33560	332	5	google	google	PROPN
bracis-33560	332	6	scholar	scholar	NOUN
bracis-33560	332	7	  	  	SPACE
bracis-33560	332	8	rae	rae	PROPN
bracis-33560	332	9	,	,	PUNCT
bracis-33560	332	10	j.w	j.w	PROPN
bracis-33560	332	11	.	.	PROPN
bracis-33560	332	12	,	,	PUNCT
bracis-33560	332	13	et	et	PROPN
bracis-33560	332	14	 	 	SPACE
bracis-33560	332	15	al	al	PROPN
bracis-33560	332	16	.	.	PROPN
bracis-33560	332	17	:	:	PUNCT
bracis-33560	332	18	scaling	scale	VERB
bracis-33560	332	19	language	language	NOUN
bracis-33560	332	20	models	model	NOUN
bracis-33560	332	21	:	:	PUNCT
bracis-33560	332	22	methods	method	NOUN
bracis-33560	332	23	,	,	PUNCT
bracis-33560	332	24	analysis	analysis	NOUN
bracis-33560	332	25	&	&	CCONJ
bracis-33560	332	26	insights	insight	NOUN
bracis-33560	332	27	from	from	ADP
bracis-33560	332	28	training	train	VERB
bracis-33560	332	29	gopher	gopher	NOUN
bracis-33560	332	30	,	,	PUNCT
bracis-33560	332	31	december	december	PROPN
bracis-33560	332	32	2021	2021	NUM
bracis-33560	332	33	google	google	PROPN
bracis-33560	332	34	scholar	scholar	NOUN
bracis-33560	332	35	  	  	SPACE
bracis-33560	332	36	raffel	raffel	NOUN
bracis-33560	332	37	,	,	PUNCT
bracis-33560	332	38	c.	c.	PROPN
bracis-33560	332	39	,	,	PUNCT
bracis-33560	332	40	et	et	PROPN
bracis-33560	332	41	 	 	SPACE
bracis-33560	332	42	al	al	PROPN
bracis-33560	332	43	.	.	PROPN
bracis-33560	332	44	:	:	PUNCT
bracis-33560	332	45	exploring	explore	VERB
bracis-33560	332	46	the	the	DET
bracis-33560	332	47	limits	limit	NOUN
bracis-33560	332	48	of	of	ADP
bracis-33560	332	49	transfer	transfer	NOUN
bracis-33560	332	50	learning	learn	VERB
bracis-33560	332	51	with	with	ADP
bracis-33560	332	52	a	a	DET
bracis-33560	332	53	unified	unified	ADJ
bracis-33560	332	54	text	text	NOUN
bracis-33560	332	55	-	-	PUNCT
bracis-33560	332	56	to	to	ADP
bracis-33560	332	57	-	-	PUNCT
bracis-33560	332	58	text	text	NOUN
bracis-33560	332	59	transformer	transformer	NOUN
bracis-33560	332	60	.	.	PUNCT
bracis-33560	333	1	j.	j.	PROPN
bracis-33560	333	2	mach	mach	PROPN
bracis-33560	333	3	.	.	PUNCT
bracis-33560	334	1	learn	learn	VERB
bracis-33560	334	2	.	.	PUNCT
bracis-33560	335	1	res	re	NOUN
bracis-33560	335	2	.	.	PROPN
bracis-33560	336	1	21(140	21(140	NUM
bracis-33560	336	2	)	)	PUNCT
bracis-33560	336	3	,	,	PUNCT
bracis-33560	336	4	1–67	1–67	PROPN
bracis-33560	336	5	(	(	PUNCT
bracis-33560	336	6	2020	2020	NUM
bracis-33560	336	7	)	)	PUNCT
bracis-33560	336	8	.	.	PUNCT
bracis-33560	337	1	http://jmlr.org/papers/v21/20-074.html	http://jmlr.org/papers/v21/20-074.html	PROPN
bracis-33560	337	2	richards	richards	PROPN
bracis-33560	337	3	,	,	PUNCT
bracis-33560	337	4	b.	b.	PROPN
bracis-33560	337	5	:	:	PUNCT
bracis-33560	337	6	type	type	NOUN
bracis-33560	337	7	/	/	SYM
bracis-33560	337	8	token	token	ADJ
bracis-33560	337	9	ratios	ratio	NOUN
bracis-33560	337	10	:	:	PUNCT
bracis-33560	337	11	what	what	PRON
bracis-33560	337	12	do	do	AUX
bracis-33560	337	13	they	they	PRON
bracis-33560	337	14	really	really	ADV
bracis-33560	337	15	tell	tell	VERB
bracis-33560	337	16	us	we	PRON
bracis-33560	337	17	?	?	PUNCT
bracis-33560	338	1	j.	j.	PROPN
bracis-33560	338	2	child	child	PROPN
bracis-33560	338	3	lang	lang	PROPN
bracis-33560	338	4	.	.	PUNCT
bracis-33560	339	1	14(2	14(2	NUM
bracis-33560	339	2	)	)	PUNCT
bracis-33560	339	3	,	,	PUNCT
bracis-33560	340	1	201–209	201–209	NUM
bracis-33560	340	2	(	(	PUNCT
bracis-33560	340	3	1987	1987	NUM
bracis-33560	340	4	)	)	PUNCT
bracis-33560	340	5	.	.	PUNCT
bracis-33560	341	1	https://doi.org/10.1017/s0305000900012885	https://doi.org/10.1017/s0305000900012885	NUM
bracis-33560	341	2	article	article	NOUN
bracis-33560	341	3	  	  	SPACE
bracis-33560	341	4	mathscinet	mathscinet	NOUN
bracis-33560	341	5	  	  	SPACE
bracis-33560	341	6	math	math	NOUN
bracis-33560	341	7	  	  	SPACE
bracis-33560	341	8	google	google	PROPN
bracis-33560	341	9	scholar	scholar	NOUN
bracis-33560	341	10	  	  	SPACE
bracis-33560	341	11	shin	shin	NOUN
bracis-33560	341	12	,	,	PUNCT
bracis-33560	341	13	s.	s.	PROPN
bracis-33560	341	14	,	,	PUNCT
bracis-33560	341	15	et	et	PROPN
bracis-33560	341	16	al	al	PROPN
bracis-33560	341	17	.	.	PUNCT
bracis-33560	341	18	:	:	PUNCT
bracis-33560	341	19	on	on	ADP
bracis-33560	341	20	the	the	DET
bracis-33560	341	21	effect	effect	NOUN
bracis-33560	341	22	of	of	ADP
bracis-33560	341	23	pretraining	pretraine	VERB
bracis-33560	341	24	corpora	corpora	NOUN
bracis-33560	341	25	on	on	ADP
bracis-33560	341	26	in	in	ADP
bracis-33560	341	27	-	-	PUNCT
bracis-33560	341	28	context	context	NOUN
bracis-33560	341	29	learning	learning	NOUN
bracis-33560	341	30	by	by	ADP
bracis-33560	341	31	a	a	DET
bracis-33560	341	32	large	large	ADJ
bracis-33560	341	33	-	-	PUNCT
bracis-33560	341	34	scale	scale	NOUN
bracis-33560	341	35	language	language	NOUN
bracis-33560	341	36	model	model	NOUN
bracis-33560	341	37	(	(	PUNCT
bracis-33560	341	38	2022	2022	NUM
bracis-33560	341	39	)	)	PUNCT
bracis-33560	341	40	google	google	PROPN
bracis-33560	341	41	scholar	scholar	NOUN
bracis-33560	341	42	  	  	SPACE
bracis-33560	341	43	souza	souza	PROPN
bracis-33560	341	44	,	,	PUNCT
bracis-33560	341	45	f.	f.	PROPN
bracis-33560	341	46	,	,	PUNCT
bracis-33560	341	47	nogueira	nogueira	PROPN
bracis-33560	341	48	,	,	PUNCT
bracis-33560	341	49	r.	r.	PROPN
bracis-33560	341	50	,	,	PUNCT
bracis-33560	341	51	lotufo	lotufo	PROPN
bracis-33560	341	52	,	,	PUNCT
bracis-33560	341	53	r.	r.	PROPN
bracis-33560	341	54	:	:	PUNCT
bracis-33560	341	55	bertimbau	bertimbau	PROPN
bracis-33560	341	56	:	:	PUNCT
bracis-33560	341	57	pretrained	pretraine	VERB
bracis-33560	341	58	bert	bert	NOUN
bracis-33560	341	59	models	model	NOUN
bracis-33560	341	60	for	for	ADP
bracis-33560	341	61	brazilian	brazilian	ADJ
bracis-33560	341	62	portuguese	portuguese	NOUN
bracis-33560	341	63	.	.	PUNCT
bracis-33560	342	1	in	in	ADP
bracis-33560	342	2	:	:	PUNCT
bracis-33560	342	3	cerri	cerri	PROPN
bracis-33560	342	4	,	,	PUNCT
bracis-33560	342	5	r.	r.	PROPN
bracis-33560	342	6	,	,	PUNCT
bracis-33560	342	7	prati	prati	PROPN
bracis-33560	342	8	,	,	PUNCT
bracis-33560	342	9	r.c	r.c	PROPN
bracis-33560	342	10	.	.	PROPN
bracis-33560	342	11	(	(	PUNCT
bracis-33560	342	12	eds	ed	NOUN
bracis-33560	342	13	.	.	PUNCT
bracis-33560	342	14	)	)	PUNCT
bracis-33560	342	15	bracis	bracis	PROPN
bracis-33560	342	16	2020	2020	NUM
bracis-33560	342	17	.	.	PUNCT
bracis-33560	343	1	lncs	lncs	PROPN
bracis-33560	343	2	(	(	PUNCT
bracis-33560	343	3	lnai	lnai	ADJ
bracis-33560	343	4	)	)	PUNCT
bracis-33560	343	5	,	,	PUNCT
bracis-33560	343	6	vol	vol	NOUN
bracis-33560	343	7	.	.	PUNCT
bracis-33560	343	8	12319	12319	NUM
bracis-33560	343	9	,	,	PUNCT
bracis-33560	343	10	pp	pp	ADP
bracis-33560	343	11	.	.	PUNCT
bracis-33560	344	1	403–417	403–417	NUM
bracis-33560	344	2	.	.	PUNCT
bracis-33560	344	3	springer	springer	NOUN
bracis-33560	344	4	,	,	PUNCT
bracis-33560	344	5	cham	cham	PROPN
bracis-33560	344	6	(	(	PUNCT
bracis-33560	344	7	2020	2020	NUM
bracis-33560	344	8	)	)	PUNCT
bracis-33560	344	9	.	.	PUNCT
bracis-33560	345	1	https://doi.org/10.1007/978-3-030-61377-8_28	https://doi.org/10.1007/978-3-030-61377-8_28	PROPN
bracis-33560	345	2	chapter	chapter	NOUN
bracis-33560	345	3	  	  	SPACE
bracis-33560	345	4	google	google	PROPN
bracis-33560	345	5	scholar	scholar	NOUN
bracis-33560	345	6	  	  	SPACE
bracis-33560	345	7	virtanen	virtanen	NOUN
bracis-33560	345	8	,	,	PUNCT
bracis-33560	345	9	a.	a.	NOUN
bracis-33560	345	10	,	,	PUNCT
bracis-33560	345	11	et	et	PROPN
bracis-33560	345	12	 	 	SPACE
bracis-33560	345	13	al	al	PROPN
bracis-33560	345	14	.	.	PROPN
bracis-33560	345	15	:	:	PUNCT
bracis-33560	346	1	multilingual	multilingual	NOUN
bracis-33560	346	2	is	be	AUX
bracis-33560	346	3	not	not	PART
bracis-33560	346	4	enough	enough	ADJ
bracis-33560	346	5	:	:	PUNCT
bracis-33560	346	6	bert	bert	PROPN
bracis-33560	346	7	for	for	ADP
bracis-33560	346	8	finnish	finnish	ADJ
bracis-33560	346	9	,	,	PUNCT
bracis-33560	346	10	december	december	PROPN
bracis-33560	346	11	2019	2019	NUM
bracis-33560	346	12	google	google	PROPN
bracis-33560	346	13	scholar	scholar	NOUN
bracis-33560	346	14	  	  	SPACE
bracis-33560	346	15	wagner	wagner	NOUN
bracis-33560	346	16	 	 	SPACE
bracis-33560	346	17	filho	filho	PROPN
bracis-33560	346	18	,	,	PUNCT
bracis-33560	346	19	j.a	j.a	PROPN
bracis-33560	346	20	.	.	PROPN
bracis-33560	346	21	,	,	PUNCT
bracis-33560	346	22	et	et	PROPN
bracis-33560	346	23	 	 	SPACE
bracis-33560	346	24	al	al	PROPN
bracis-33560	346	25	.	.	PROPN
bracis-33560	346	26	:	:	PUNCT
bracis-33560	347	1	the	the	DET
bracis-33560	347	2	brwac	brwac	PROPN
bracis-33560	347	3	corpus	corpus	PROPN
bracis-33560	347	4	:	:	PUNCT
bracis-33560	347	5	a	a	DET
bracis-33560	347	6	new	new	ADJ
bracis-33560	347	7	open	open	ADJ
bracis-33560	347	8	resource	resource	NOUN
bracis-33560	347	9	for	for	ADP
bracis-33560	347	10	brazilian	brazilian	ADJ
bracis-33560	347	11	portuguese	portuguese	NOUN
bracis-33560	347	12	.	.	PUNCT
bracis-33560	348	1	in	in	ADP
bracis-33560	348	2	:	:	PUNCT
bracis-33560	348	3	calzolari	calzolari	PROPN
bracis-33560	348	4	,	,	PUNCT
bracis-33560	348	5	n.	n.	PROPN
bracis-33560	348	6	,	,	PUNCT
bracis-33560	348	7	et	et	PROPN
bracis-33560	348	8	 	 	SPACE
bracis-33560	348	9	al	al	PROPN
bracis-33560	348	10	.	.	PROPN
bracis-33560	348	11	(	(	PUNCT
bracis-33560	348	12	eds	ed	NOUN
bracis-33560	348	13	.	.	PUNCT
bracis-33560	348	14	)	)	PUNCT
bracis-33560	348	15	proceedings	proceeding	NOUN
bracis-33560	348	16	of	of	ADP
bracis-33560	348	17	the	the	DET
bracis-33560	348	18	eleventh	eleventh	ADJ
bracis-33560	348	19	international	international	ADJ
bracis-33560	348	20	conference	conference	NOUN
bracis-33560	348	21	on	on	ADP
bracis-33560	348	22	language	language	NOUN
bracis-33560	348	23	resources	resource	NOUN
bracis-33560	348	24	and	and	CCONJ
bracis-33560	348	25	evaluation	evaluation	NOUN
bracis-33560	348	26	(	(	PUNCT
bracis-33560	348	27	lrec	lrec	PROPN
bracis-33560	348	28	2018	2018	NUM
bracis-33560	348	29	)	)	PUNCT
bracis-33560	348	30	.	.	PUNCT
bracis-33560	349	1	european	european	ADJ
bracis-33560	349	2	language	language	PROPN
bracis-33560	349	3	resources	resource	NOUN
bracis-33560	349	4	association	association	NOUN
bracis-33560	349	5	(	(	PUNCT
bracis-33560	349	6	elra	elra	ADV
bracis-33560	349	7	)	)	PUNCT
bracis-33560	349	8	,	,	PUNCT
bracis-33560	349	9	miyazaki	miyazaki	PROPN
bracis-33560	349	10	,	,	PUNCT
bracis-33560	349	11	japan	japan	PROPN
bracis-33560	349	12	,	,	PUNCT
bracis-33560	349	13	may	may	PROPN
bracis-33560	349	14	2018	2018	NUM
bracis-33560	349	15	.	.	PUNCT
bracis-33560	350	1	https://aclanthology.org/l18-1686	https://aclanthology.org/l18-1686	PROPN
bracis-33560	351	1	xu	xu	PROPN
bracis-33560	351	2	,	,	PUNCT
bracis-33560	351	3	l.	l.	PROPN
bracis-33560	351	4	,	,	PUNCT
bracis-33560	351	5	zhang	zhang	PROPN
bracis-33560	351	6	,	,	PUNCT
bracis-33560	351	7	x.	x.	PROPN
bracis-33560	351	8	,	,	PUNCT
bracis-33560	351	9	dong	dong	PROPN
bracis-33560	351	10	,	,	PUNCT
bracis-33560	351	11	q.	q.	PROPN
bracis-33560	351	12	:	:	PUNCT
bracis-33560	352	1	cluecorpus2020	cluecorpus2020	ADP
bracis-33560	352	2	:	:	PUNCT
bracis-33560	352	3	a	a	DET
bracis-33560	352	4	large	large	ADJ
bracis-33560	352	5	-	-	PUNCT
bracis-33560	352	6	scale	scale	NOUN
bracis-33560	352	7	chinese	chinese	ADJ
bracis-33560	352	8	corpus	corpus	NOUN
bracis-33560	352	9	for	for	ADP
bracis-33560	352	10	pre	pre	ADJ
bracis-33560	352	11	-	-	ADJ
bracis-33560	352	12	training	training	ADJ
bracis-33560	352	13	language	language	NOUN
bracis-33560	352	14	model	model	NOUN
bracis-33560	352	15	.	.	PUNCT
bracis-33560	353	1	arxiv	arxiv	PROPN
bracis-33560	353	2	preprint	preprint	VERB
bracis-33560	353	3	arxiv:2003.01355	arxiv:2003.01355	NUM
bracis-33560	353	4	(	(	PUNCT
bracis-33560	353	5	2020	2020	NUM
bracis-33560	353	6	)	)	PUNCT
bracis-33560	353	7	youmans	youman	NOUN
bracis-33560	353	8	,	,	PUNCT
bracis-33560	353	9	g.	g.	NOUN
bracis-33560	353	10	:	:	PUNCT
bracis-33560	353	11	measuring	measure	VERB
bracis-33560	353	12	lexical	lexical	ADJ
bracis-33560	353	13	style	style	NOUN
bracis-33560	353	14	and	and	CCONJ
bracis-33560	353	15	competence	competence	NOUN
bracis-33560	353	16	:	:	PUNCT
bracis-33560	353	17	the	the	DET
bracis-33560	353	18	type	type	NOUN
bracis-33560	353	19	-	-	PUNCT
bracis-33560	353	20	token	token	VERB
bracis-33560	353	21	vocabulary	vocabulary	ADJ
bracis-33560	353	22	curve	curve	NOUN
bracis-33560	353	23	.	.	PUNCT
bracis-33560	354	1	style	style	NOUN
bracis-33560	354	2	24(4	24(4	NOUN
bracis-33560	354	3	)	)	PUNCT
bracis-33560	354	4	,	,	PUNCT
bracis-33560	354	5	584–599	584–599	NUM
bracis-33560	354	6	(	(	PUNCT
bracis-33560	354	7	1990	1990	NUM
bracis-33560	354	8	)	)	PUNCT
bracis-33560	354	9	.	.	PUNCT
bracis-33560	355	1	http://www.jstor.org/stable/42946163	http://www.jstor.org/stable/42946163	PROPN
bracis-33560	355	2	yuan	yuan	PROPN
bracis-33560	355	3	,	,	PUNCT
bracis-33560	355	4	s.	s.	PROPN
bracis-33560	355	5	,	,	PUNCT
bracis-33560	355	6	et	et	PROPN
bracis-33560	355	7	al	al	PROPN
bracis-33560	355	8	.	.	PUNCT
bracis-33560	355	9	:	:	PUNCT
bracis-33560	356	1	wudaocorpora	wudaocorpora	INTJ
bracis-33560	356	2	:	:	PUNCT
bracis-33560	356	3	a	a	DET
bracis-33560	356	4	super	super	ADV
bracis-33560	356	5	large	large	ADJ
bracis-33560	356	6	-	-	PUNCT
bracis-33560	356	7	scale	scale	NOUN
bracis-33560	356	8	chinese	chinese	PROPN
bracis-33560	356	9	corpora	corpora	PROPN
bracis-33560	356	10	for	for	ADP
bracis-33560	356	11	pre	pre	ADJ
bracis-33560	356	12	-	-	ADJ
bracis-33560	356	13	training	training	ADJ
bracis-33560	356	14	language	language	NOUN
bracis-33560	356	15	models	model	NOUN
bracis-33560	356	16	.	.	PUNCT
bracis-33560	357	1	ai	ai	VERB
bracis-33560	357	2	open	open	ADJ
bracis-33560	357	3	2	2	NUM
bracis-33560	357	4	,	,	PUNCT
bracis-33560	357	5	65–68	65–68	NUM
bracis-33560	357	6	(	(	PUNCT
bracis-33560	357	7	2021	2021	NUM
bracis-33560	357	8	)	)	PUNCT
bracis-33560	357	9	article	article	NOUN
bracis-33560	357	10	  	  	SPACE
bracis-33560	357	11	math	math	NOUN
bracis-33560	357	12	  	  	SPACE
bracis-33560	357	13	google	google	PROPN
bracis-33560	357	14	scholar	scholar	NOUN
bracis-33560	357	15	  	  	SPACE
bracis-33560	357	16	download	download	NOUN
bracis-33560	357	17	references	reference	NOUN
bracis-33560	357	18	acknowledgements	acknowledgement	NOUN
bracis-33560	357	19	we	we	PRON
bracis-33560	357	20	thank	thank	VERB
bracis-33560	357	21	instituto	instituto	PROPN
bracis-33560	357	22	de	de	PROPN
bracis-33560	357	23	ciência	ciência	PROPN
bracis-33560	357	24	e	e	PROPN
bracis-33560	357	25	tecnologia	tecnologia	PROPN
bracis-33560	357	26	itaú	itaú	NOUN
bracis-33560	357	27	-	-	PUNCT
bracis-33560	357	28	unibanco	unibanco	NOUN
bracis-33560	357	29	(	(	PUNCT
bracis-33560	357	30	icti	icti	ADJ
bracis-33560	357	31	)	)	PUNCT
bracis-33560	357	32	and	and	CCONJ
bracis-33560	357	33	itaú	itaú	NOUN
bracis-33560	357	34	-	-	PUNCT
bracis-33560	357	35	unibanco	unibanco	NOUN
bracis-33560	357	36	sa	sa	NOUN
bracis-33560	357	37	for	for	ADP
bracis-33560	357	38	the	the	DET
bracis-33560	357	39	technical	technical	ADJ
bracis-33560	357	40	support	support	NOUN
bracis-33560	357	41	,	,	PUNCT
bracis-33560	357	42	resources	resource	NOUN
bracis-33560	357	43	,	,	PUNCT
bracis-33560	357	44	and	and	CCONJ
bracis-33560	357	45	financial	financial	ADJ
bracis-33560	357	46	aid	aid	NOUN
bracis-33560	357	47	in	in	ADP
bracis-33560	357	48	the	the	DET
bracis-33560	357	49	development	development	NOUN
bracis-33560	357	50	of	of	ADP
bracis-33560	357	51	the	the	DET
bracis-33560	357	52	aroeira	aroeira	PROPN
bracis-33560	357	53	corpus	corpus	PROPN
bracis-33560	357	54	.	.	PUNCT
bracis-33560	358	1	it	it	PRON
bracis-33560	358	2	’s	’	VERB
bracis-33560	358	3	also	also	ADV
bracis-33560	358	4	noteworthy	noteworthy	ADJ
bracis-33560	358	5	the	the	DET
bracis-33560	358	6	fact	fact	NOUN
bracis-33560	358	7	that	that	SCONJ
bracis-33560	358	8	chatgpt	chatgpt	NOUN
bracis-33560	358	9	(	(	PUNCT
bracis-33560	358	10	openai	openai	PROPN
bracis-33560	358	11	)	)	PUNCT
bracis-33560	358	12	was	be	AUX
bracis-33560	358	13	employed	employ	VERB
bracis-33560	358	14	in	in	ADP
bracis-33560	358	15	the	the	DET
bracis-33560	358	16	writing	writing	NOUN
bracis-33560	358	17	process	process	NOUN
bracis-33560	358	18	,	,	PUNCT
bracis-33560	358	19	contributing	contribute	VERB
bracis-33560	358	20	to	to	ADP
bracis-33560	358	21	thorough	thorough	ADJ
bracis-33560	358	22	grammatical	grammatical	ADJ
bracis-33560	358	23	and	and	CCONJ
bracis-33560	358	24	semantic	semantic	ADJ
bracis-33560	358	25	reviews	review	NOUN
bracis-33560	358	26	.	.	PUNCT
bracis-33560	359	1	author	author	NOUN
bracis-33560	359	2	information	information	NOUN
bracis-33560	359	3	authors	author	NOUN
bracis-33560	359	4	and	and	CCONJ
bracis-33560	359	5	affiliations	affiliation	NOUN
bracis-33560	359	6	instituto	instituto	PROPN
bracis-33560	359	7	de	de	PROPN
bracis-33560	359	8	ciência	ciência	PROPN
bracis-33560	359	9	e	e	PROPN
bracis-33560	359	10	tecnologia	tecnologia	PROPN
bracis-33560	359	11	itaú	itaú	NOUN
bracis-33560	359	12	-	-	PUNCT
bracis-33560	359	13	unibanco	unibanco	NOUN
bracis-33560	359	14	(	(	PUNCT
bracis-33560	359	15	icti	icti	ADJ
bracis-33560	359	16	)	)	PUNCT
bracis-33560	359	17	,	,	PUNCT
bracis-33560	359	18	são	são	PROPN
bracis-33560	359	19	paulo	paulo	PROPN
bracis-33560	359	20	,	,	PUNCT
bracis-33560	359	21	brazil	brazil	PROPN
bracis-33560	359	22	thiago	thiago	PROPN
bracis-33560	359	23	lira	lira	PROPN
bracis-33560	359	24	,	,	PUNCT
bracis-33560	359	25	 	 	SPACE
bracis-33560	359	26	flávio	flávio	NOUN
bracis-33560	359	27	cação	cação	NOUN
bracis-33560	359	28	,	,	PUNCT
bracis-33560	359	29	 	 	SPACE
bracis-33560	359	30	cinthia	cinthia	NOUN
bracis-33560	359	31	souza	souza	PROPN
bracis-33560	359	32	,	,	PUNCT
bracis-33560	359	33	 	 	SPACE
bracis-33560	359	34	joão	joão	PROPN
bracis-33560	359	35	valentini	valentini	PROPN
bracis-33560	359	36	,	,	PUNCT
bracis-33560	359	37	 	 	SPACE
bracis-33560	359	38	edson	edson	PROPN
bracis-33560	359	39	bollis	bollis	PROPN
bracis-33560	359	40	,	,	PUNCT
bracis-33560	359	41	 	 	SPACE
bracis-33560	359	42	otavio	otavio	NOUN
bracis-33560	359	43	oliveira	oliveira	PROPN
bracis-33560	359	44	,	,	PUNCT
bracis-33560	359	45	 	 	SPACE
bracis-33560	359	46	renato	renato	PROPN
bracis-33560	359	47	almeida	almeida	PROPN
bracis-33560	359	48	,	,	PUNCT
bracis-33560	359	49	 	 	SPACE
bracis-33560	359	50	marcio	marcio	PROPN
bracis-33560	359	51	magalhães	magalhães	NOUN
bracis-33560	359	52	,	,	PUNCT
bracis-33560	359	53	 	 	SPACE
bracis-33560	359	54	katia	katia	PROPN
bracis-33560	359	55	poloni	poloni	PROPN
bracis-33560	359	56	,	,	PUNCT
bracis-33560	359	57	 	 	SPACE
bracis-33560	359	58	andre	andre	PROPN
bracis-33560	359	59	oliveira	oliveira	PROPN
bracis-33560	359	60	 	 	SPACE
bracis-33560	359	61	&	&	CCONJ
bracis-33560	359	62	 	 	SPACE
bracis-33560	359	63	lucas	lucas	PROPN
bracis-33560	359	64	pellicer	pellicer	PROPN
bracis-33560	359	65	authors	author	NOUN
bracis-33560	359	66	thiago	thiago	PROPN
bracis-33560	359	67	liraview	liraview	PROPN
bracis-33560	359	68	author	author	NOUN
bracis-33560	359	69	publications	publication	NOUN
bracis-33560	359	70	search	search	NOUN
bracis-33560	359	71	author	author	NOUN
bracis-33560	359	72	on	on	ADP
bracis-33560	359	73	:	:	PUNCT
bracis-33560	359	74	pubmed	pubmed	PROPN
bracis-33560	359	75	 	 	SPACE
bracis-33560	359	76	google	google	PROPN
bracis-33560	359	77	scholar	scholar	NOUN
bracis-33560	359	78	flávio	flávio	NOUN
bracis-33560	359	79	caçãoview	caçãoview	NOUN
bracis-33560	359	80	author	author	NOUN
bracis-33560	359	81	publications	publication	NOUN
bracis-33560	359	82	search	search	NOUN
bracis-33560	359	83	author	author	NOUN
bracis-33560	359	84	on	on	ADP
bracis-33560	359	85	:	:	PUNCT
bracis-33560	359	86	pubmed	pubmed	PROPN
bracis-33560	359	87	 	 	SPACE
bracis-33560	359	88	google	google	PROPN
bracis-33560	359	89	scholar	scholar	NOUN
bracis-33560	359	90	cinthia	cinthia	NOUN
bracis-33560	359	91	souzaview	souzaview	NOUN
bracis-33560	359	92	author	author	NOUN
bracis-33560	359	93	publications	publication	NOUN
bracis-33560	359	94	search	search	NOUN
bracis-33560	359	95	author	author	NOUN
bracis-33560	359	96	on	on	ADP
bracis-33560	359	97	:	:	PUNCT
bracis-33560	359	98	pubmed	pubmed	PROPN
bracis-33560	359	99	 	 	SPACE
bracis-33560	359	100	google	google	PROPN
bracis-33560	359	101	scholar	scholar	PROPN
bracis-33560	359	102	joão	joão	PROPN
bracis-33560	359	103	valentiniview	valentiniview	PROPN
bracis-33560	359	104	author	author	NOUN
bracis-33560	359	105	publications	publication	NOUN
bracis-33560	359	106	search	search	NOUN
bracis-33560	359	107	author	author	NOUN
bracis-33560	359	108	on	on	ADP
bracis-33560	359	109	:	:	PUNCT
bracis-33560	359	110	pubmed	pubmed	PROPN
bracis-33560	359	111	 	 	SPACE
bracis-33560	359	112	google	google	PROPN
bracis-33560	359	113	scholar	scholar	PROPN
bracis-33560	359	114	edson	edson	PROPN
bracis-33560	359	115	bollisview	bollisview	PROPN
bracis-33560	359	116	author	author	NOUN
bracis-33560	359	117	publications	publication	NOUN
bracis-33560	359	118	search	search	NOUN
bracis-33560	359	119	author	author	NOUN
bracis-33560	359	120	on	on	ADP
bracis-33560	359	121	:	:	PUNCT
bracis-33560	359	122	pubmed	pubmed	PROPN
bracis-33560	359	123	 	 	SPACE
bracis-33560	359	124	google	google	PROPN
bracis-33560	359	125	scholar	scholar	NOUN
bracis-33560	359	126	otavio	otavio	NOUN
bracis-33560	359	127	oliveiraview	oliveiraview	NOUN
bracis-33560	359	128	author	author	NOUN
bracis-33560	359	129	publications	publication	NOUN
bracis-33560	359	130	search	search	NOUN
bracis-33560	359	131	author	author	NOUN
bracis-33560	359	132	on	on	ADP
bracis-33560	359	133	:	:	PUNCT
bracis-33560	359	134	pubmed	pubmed	PROPN
bracis-33560	359	135	 	 	SPACE
bracis-33560	359	136	google	google	PROPN
bracis-33560	359	137	scholar	scholar	PROPN
bracis-33560	359	138	renato	renato	PROPN
bracis-33560	359	139	almeidaview	almeidaview	NOUN
bracis-33560	359	140	author	author	NOUN
bracis-33560	359	141	publications	publication	VERB
bracis-33560	359	142	search	search	NOUN
bracis-33560	359	143	author	author	NOUN
bracis-33560	359	144	on	on	ADP
bracis-33560	359	145	:	:	PUNCT
bracis-33560	359	146	pubmed	pubmed	PROPN
bracis-33560	359	147	 	 	SPACE
bracis-33560	359	148	google	google	PROPN
bracis-33560	359	149	scholar	scholar	PROPN
bracis-33560	359	150	marcio	marcio	PROPN
bracis-33560	359	151	magalhãesview	magalhãesview	PROPN
bracis-33560	359	152	author	author	NOUN
bracis-33560	359	153	publications	publication	VERB
bracis-33560	359	154	search	search	NOUN
bracis-33560	359	155	author	author	NOUN
bracis-33560	359	156	on	on	ADP
bracis-33560	359	157	:	:	PUNCT
bracis-33560	359	158	pubmed	pubmed	PROPN
bracis-33560	359	159	 	 	SPACE
bracis-33560	359	160	google	google	PROPN
bracis-33560	359	161	scholar	scholar	PROPN
bracis-33560	359	162	katia	katia	PROPN
bracis-33560	359	163	poloniview	poloniview	PROPN
bracis-33560	359	164	author	author	NOUN
bracis-33560	359	165	publications	publication	NOUN
bracis-33560	359	166	search	search	NOUN
bracis-33560	359	167	author	author	NOUN
bracis-33560	359	168	on	on	ADP
bracis-33560	359	169	:	:	PUNCT
bracis-33560	359	170	pubmed	pubmed	PROPN
bracis-33560	359	171	 	 	SPACE
bracis-33560	359	172	google	google	PROPN
bracis-33560	359	173	scholar	scholar	NOUN
bracis-33560	359	174	andre	andre	PROPN
bracis-33560	359	175	oliveiraview	oliveiraview	PROPN
bracis-33560	359	176	author	author	NOUN
bracis-33560	359	177	publications	publication	NOUN
bracis-33560	359	178	search	search	NOUN
bracis-33560	359	179	author	author	NOUN
bracis-33560	359	180	on	on	ADP
bracis-33560	359	181	:	:	PUNCT
bracis-33560	359	182	pubmed	pubmed	PROPN
bracis-33560	359	183	 	 	SPACE
bracis-33560	359	184	google	google	PROPN
bracis-33560	359	185	scholar	scholar	NOUN
bracis-33560	359	186	lucas	lucas	PROPN
bracis-33560	359	187	pellicerview	pellicerview	NOUN
bracis-33560	359	188	author	author	NOUN
bracis-33560	359	189	publications	publication	NOUN
bracis-33560	359	190	search	search	NOUN
bracis-33560	359	191	author	author	NOUN
bracis-33560	359	192	on	on	ADP
bracis-33560	359	193	:	:	PUNCT
bracis-33560	359	194	pubmed	pubmed	PROPN
bracis-33560	359	195	 	 	SPACE
bracis-33560	359	196	google	google	PROPN
bracis-33560	359	197	scholar	scholar	NOUN
bracis-33560	359	198	corresponding	correspond	VERB
bracis-33560	359	199	author	author	NOUN
bracis-33560	359	200	correspondence	correspondence	NOUN
bracis-33560	359	201	to	to	ADP
bracis-33560	359	202	lucas	lucas	PROPN
bracis-33560	359	203	pellicer	pellicer	PROPN
bracis-33560	359	204	.	.	PUNCT
bracis-33560	360	1	editor	editor	NOUN
bracis-33560	360	2	information	information	NOUN
bracis-33560	360	3	editors	editor	NOUN
bracis-33560	360	4	and	and	CCONJ
bracis-33560	360	5	affiliations	affiliation	NOUN
bracis-33560	360	6	universidade	universidade	VERB
bracis-33560	360	7	federal	federal	PROPN
bracis-33560	360	8	fluminense	fluminense	PROPN
bracis-33560	360	9	,	,	PUNCT
bracis-33560	360	10	niterói	niterói	PROPN
bracis-33560	360	11	,	,	PUNCT
bracis-33560	360	12	brazil	brazil	PROPN
bracis-33560	360	13	aline	aline	PROPN
bracis-33560	360	14	paes	paes	PROPN
bracis-33560	360	15	instituto	instituto	PROPN
bracis-33560	360	16	tecnológico	tecnológico	PROPN
bracis-33560	360	17	de	de	PROPN
bracis-33560	360	18	aeronáutica	aeronáutica	PROPN
bracis-33560	360	19	,	,	PUNCT
bracis-33560	360	20	são	são	PROPN
bracis-33560	360	21	josé	josé	PROPN
bracis-33560	360	22	dos	dos	PROPN
bracis-33560	360	23	campos	campos	PROPN
bracis-33560	360	24	,	,	PUNCT
bracis-33560	361	1	brazil	brazil	PROPN
bracis-33560	361	2	filipe	filipe	PROPN
bracis-33560	361	3	a.	a.	PROPN
bracis-33560	361	4	n.	n.	PROPN
bracis-33560	361	5	verri	verri	PROPN
bracis-33560	361	6	ethics	ethic	NOUN
bracis-33560	361	7	declarations	declaration	NOUN
bracis-33560	361	8	data	datum	NOUN
bracis-33560	361	9	availability	availability	NOUN
bracis-33560	361	10	our	our	PRON
bracis-33560	361	11	aroeira	aroeira	PROPN
bracis-33560	361	12	corpus	corpus	PROPN
bracis-33560	361	13	is	be	AUX
bracis-33560	361	14	available	available	ADJ
bracis-33560	361	15	for	for	ADP
bracis-33560	361	16	download	download	NOUN
bracis-33560	361	17	in	in	ADP
bracis-33560	361	18	the	the	DET
bracis-33560	361	19	hugging	hugging	NOUN
bracis-33560	361	20	face	face	NOUN
bracis-33560	361	21	repository	repository	NOUN
bracis-33560	361	22	:	:	PUNCT
bracis-33560	361	23	https://huggingface.co/datasets/itau-unibanco/aroeira	https://huggingface.co/datasets/itau-unibanco/aroeira	PROPN
bracis-33560	361	24	and	and	CCONJ
bracis-33560	361	25	is	be	AUX
bracis-33560	361	26	under	under	ADP
bracis-33560	361	27	the	the	DET
bracis-33560	361	28	cc	cc	NOUN
bracis-33560	361	29	-	-	PUNCT
bracis-33560	361	30	by	by	ADP
bracis-33560	361	31	-	-	PUNCT
bracis-33560	361	32	nc	nc	NOUN
bracis-33560	361	33	4.0	4.0	NUM
bracis-33560	361	34	license	license	NOUN
bracis-33560	361	35	.	.	PUNCT
bracis-33560	362	1	rights	right	NOUN
bracis-33560	362	2	and	and	CCONJ
bracis-33560	362	3	permissions	permission	NOUN
bracis-33560	362	4	reprints	reprint	NOUN
bracis-33560	362	5	and	and	CCONJ
bracis-33560	362	6	permissions	permission	VERB
bracis-33560	362	7	copyright	copyright	NOUN
bracis-33560	362	8	information	information	NOUN
bracis-33560	362	9	©	©	ADP
bracis-33560	362	10	2025	2025	NUM
bracis-33560	362	11	the	the	DET
bracis-33560	362	12	author(s	author(s	NOUN
bracis-33560	362	13	)	)	PUNCT
bracis-33560	362	14	,	,	PUNCT
bracis-33560	362	15	under	under	ADP
bracis-33560	362	16	exclusive	exclusive	ADJ
bracis-33560	362	17	license	license	NOUN
bracis-33560	362	18	to	to	ADP
bracis-33560	362	19	springer	springer	NOUN
bracis-33560	362	20	nature	nature	PROPN
bracis-33560	362	21	switzerland	switzerland	PROPN
bracis-33560	362	22	ag	ag	PROPN
bracis-33560	362	23	about	about	ADP
bracis-33560	362	24	this	this	DET
bracis-33560	362	25	paper	paper	NOUN
bracis-33560	362	26	cite	cite	VERB
bracis-33560	362	27	this	this	DET
bracis-33560	362	28	paper	paper	NOUN
bracis-33560	362	29	lira	lira	NOUN
bracis-33560	362	30	,	,	PUNCT
bracis-33560	362	31	t.	t.	PROPN
bracis-33560	362	32	et	et	PROPN
bracis-33560	362	33	al	al	PROPN
bracis-33560	362	34	.	.	PROPN
bracis-33560	363	1	(	(	PUNCT
bracis-33560	363	2	2025	2025	NUM
bracis-33560	363	3	)	)	PUNCT
bracis-33560	363	4	.	.	PUNCT
bracis-33560	364	1	aroeira	aroeira	PROPN
bracis-33560	364	2	:	:	PUNCT
bracis-33560	364	3	a	a	DET
bracis-33560	364	4	curated	curate	VERB
bracis-33560	364	5	corpus	corpus	NOUN
bracis-33560	364	6	for	for	ADP
bracis-33560	364	7	 	 	SPACE
bracis-33560	364	8	the	the	DET
bracis-33560	364	9	 	 	SPACE
bracis-33560	364	10	portuguese	portuguese	ADJ
bracis-33560	364	11	language	language	NOUN
bracis-33560	364	12	with	with	ADP
bracis-33560	364	13	 	 	SPACE
bracis-33560	364	14	a	a	DET
bracis-33560	364	15	 	 	SPACE
bracis-33560	364	16	large	large	ADJ
bracis-33560	364	17	number	number	NOUN
bracis-33560	364	18	of	of	ADP
bracis-33560	364	19	 	 	SPACE
bracis-33560	364	20	tokens	token	NOUN
bracis-33560	364	21	.	.	PUNCT
bracis-33560	365	1	in	in	ADP
bracis-33560	365	2	:	:	PUNCT
bracis-33560	365	3	paes	paes	PROPN
bracis-33560	365	4	,	,	PUNCT
bracis-33560	365	5	a.	a.	PROPN
bracis-33560	365	6	,	,	PUNCT
bracis-33560	365	7	verri	verri	NOUN
bracis-33560	365	8	,	,	PUNCT
bracis-33560	365	9	f.a.n	f.a.n	ADJ
bracis-33560	365	10	.	.	PUNCT
bracis-33560	365	11	(	(	PUNCT
bracis-33560	365	12	eds	ed	NOUN
bracis-33560	365	13	)	)	PUNCT
bracis-33560	365	14	intelligent	intelligent	ADJ
bracis-33560	365	15	systems	system	NOUN
bracis-33560	365	16	.	.	PUNCT
bracis-33560	366	1	bracis	bracis	PROPN
bracis-33560	366	2	2024	2024	NUM
bracis-33560	366	3	.	.	PUNCT
bracis-33560	367	1	lecture	lecture	NOUN
bracis-33560	367	2	notes	note	NOUN
bracis-33560	367	3	in	in	ADP
bracis-33560	367	4	computer	computer	NOUN
bracis-33560	367	5	science	science	NOUN
bracis-33560	367	6	(	(	PUNCT
bracis-33560	367	7	)	)	PUNCT
bracis-33560	367	8	,	,	PUNCT
bracis-33560	367	9	vol	vol	NOUN
bracis-33560	367	10	15412	15412	NUM
bracis-33560	367	11	.	.	PUNCT
bracis-33560	368	1	springer	springer	NOUN
bracis-33560	368	2	,	,	PUNCT
bracis-33560	368	3	cham	cham	PROPN
bracis-33560	368	4	.	.	PUNCT
bracis-33560	369	1	https://doi.org/10.1007/978-3-031-79029-4_13	https://doi.org/10.1007/978-3-031-79029-4_13	NOUN
bracis-33560	369	2	download	download	NOUN
bracis-33560	369	3	citation	citation	NOUN
bracis-33560	369	4	.ris	.ris	PUNCT
bracis-33560	370	1	.enw	.enw	PROPN
bracis-33560	370	2	.bib	.bib	PUNCT
bracis-33560	371	1	doi	doi	PROPN
bracis-33560	371	2	:	:	PUNCT
bracis-33560	371	3	https://doi.org/10.1007/978-3-031-79029-4_13	https://doi.org/10.1007/978-3-031-79029-4_13	NOUN
bracis-33560	371	4	published	publish	VERB
bracis-33560	371	5	:	:	PUNCT
bracis-33560	371	6	30	30	NUM
bracis-33560	371	7	january	january	PROPN
bracis-33560	371	8	2025	2025	NUM
bracis-33560	371	9	publisher	publisher	NOUN
bracis-33560	371	10	name	name	NOUN
bracis-33560	371	11	:	:	PUNCT
bracis-33560	371	12	springer	springer	NOUN
bracis-33560	371	13	,	,	PUNCT
bracis-33560	371	14	cham	cham	PROPN
bracis-33560	371	15	print	print	PROPN
bracis-33560	371	16	isbn	isbn	PROPN
bracis-33560	371	17	:	:	PUNCT
bracis-33560	371	18	978	978	NUM
bracis-33560	371	19	-	-	SYM
bracis-33560	371	20	3	3	NUM
bracis-33560	371	21	-	-	PUNCT
bracis-33560	371	22	031	031	NUM
bracis-33560	371	23	-	-	PUNCT
bracis-33560	371	24	79028	79028	NUM
bracis-33560	371	25	-	-	SYM
bracis-33560	371	26	7	7	NUM
bracis-33560	371	27	online	online	ADJ
bracis-33560	371	28	isbn	isbn	NOUN
bracis-33560	371	29	:	:	PUNCT
bracis-33560	371	30	978	978	NUM
bracis-33560	371	31	-	-	SYM
bracis-33560	371	32	3	3	NUM
bracis-33560	371	33	-	-	PUNCT
bracis-33560	371	34	031	031	NUM
bracis-33560	371	35	-	-	PUNCT
bracis-33560	371	36	79029	79029	NUM
bracis-33560	371	37	-	-	SYM
bracis-33560	371	38	4	4	NUM
bracis-33560	371	39	ebook	ebook	NOUN
bracis-33560	371	40	packages	package	NOUN
bracis-33560	371	41	:	:	PUNCT
bracis-33560	371	42	computer	computer	NOUN
bracis-33560	371	43	sciencecomputer	sciencecomputer	NOUN
bracis-33560	371	44	science	science	NOUN
bracis-33560	371	45	(	(	PUNCT
bracis-33560	371	46	r0	r0	NOUN
bracis-33560	371	47	)	)	PUNCT
bracis-33560	371	48	share	share	VERB
bracis-33560	371	49	this	this	DET
bracis-33560	371	50	paper	paper	NOUN
bracis-33560	371	51	anyone	anyone	PRON
bracis-33560	371	52	you	you	PRON
bracis-33560	371	53	share	share	VERB
bracis-33560	371	54	the	the	DET
bracis-33560	371	55	following	follow	VERB
bracis-33560	371	56	link	link	NOUN
bracis-33560	371	57	with	with	ADP
bracis-33560	371	58	will	will	AUX
bracis-33560	371	59	be	be	AUX
bracis-33560	371	60	able	able	ADJ
bracis-33560	371	61	to	to	PART
bracis-33560	371	62	read	read	VERB
bracis-33560	371	63	this	this	DET
bracis-33560	371	64	content	content	NOUN
bracis-33560	371	65	:	:	PUNCT
bracis-33560	371	66	get	get	VERB
bracis-33560	371	67	shareable	shareable	ADJ
bracis-33560	371	68	linksorry	linksorry	NOUN
bracis-33560	371	69	,	,	PUNCT
bracis-33560	371	70	a	a	DET
bracis-33560	371	71	shareable	shareable	ADJ
bracis-33560	371	72	link	link	NOUN
bracis-33560	371	73	is	be	AUX
bracis-33560	371	74	not	not	PART
bracis-33560	371	75	currently	currently	ADV
bracis-33560	371	76	available	available	ADJ
bracis-33560	371	77	for	for	ADP
bracis-33560	371	78	this	this	DET
bracis-33560	371	79	article	article	NOUN
bracis-33560	371	80	.	.	PUNCT
bracis-33560	372	1	copy	copy	VERB
bracis-33560	372	2	shareable	shareable	ADJ
bracis-33560	372	3	link	link	NOUN
bracis-33560	372	4	to	to	PART
bracis-33560	372	5	clipboard	clipboard	NOUN
bracis-33560	372	6	provided	provide	VERB
bracis-33560	372	7	by	by	ADP
bracis-33560	372	8	the	the	DET
bracis-33560	372	9	springer	springer	NOUN
bracis-33560	372	10	nature	nature	PROPN
bracis-33560	372	11	sharedit	sharedit	PROPN
bracis-33560	372	12	content	content	NOUN
bracis-33560	372	13	-	-	PUNCT
bracis-33560	372	14	sharing	share	VERB
bracis-33560	372	15	initiative	initiative	NOUN
bracis-33560	372	16	publish	publish	VERB
bracis-33560	372	17	with	with	ADP
bracis-33560	372	18	us	us	PROPN
bracis-33560	372	19	policies	policy	NOUN
bracis-33560	372	20	and	and	CCONJ
bracis-33560	372	21	ethics	ethic	NOUN
bracis-33560	372	22	search	search	NOUN
bracis-33560	372	23	search	search	NOUN
bracis-33560	372	24	by	by	ADP
bracis-33560	372	25	keyword	keyword	NOUN
bracis-33560	372	26	or	or	CCONJ
bracis-33560	372	27	author	author	NOUN
bracis-33560	372	28	search	search	NOUN
bracis-33560	372	29	navigation	navigation	NOUN
bracis-33560	372	30	find	find	VERB
bracis-33560	372	31	a	a	DET
bracis-33560	372	32	journal	journal	NOUN
bracis-33560	372	33	publish	publish	VERB
bracis-33560	372	34	with	with	ADP
bracis-33560	372	35	us	we	PRON
bracis-33560	372	36	track	track	VERB
bracis-33560	372	37	your	your	PRON
bracis-33560	372	38	research	research	NOUN
bracis-33560	372	39	discover	discover	VERB
bracis-33560	372	40	content	content	NOUN
bracis-33560	372	41	journals	journal	NOUN
bracis-33560	372	42	a	a	DET
bracis-33560	372	43	-	-	PUNCT
bracis-33560	372	44	z	z	NOUN
bracis-33560	372	45	books	book	NOUN
bracis-33560	372	46	a	a	DET
bracis-33560	372	47	-	-	PUNCT
bracis-33560	372	48	z	z	NOUN
bracis-33560	372	49	publish	publish	NOUN
bracis-33560	372	50	with	with	ADP
bracis-33560	372	51	us	us	PROPN
bracis-33560	372	52	journal	journal	PROPN
bracis-33560	372	53	finder	finder	PROPN
bracis-33560	372	54	publish	publish	VERB
bracis-33560	372	55	your	your	PRON
bracis-33560	372	56	research	research	NOUN
bracis-33560	372	57	language	language	NOUN
bracis-33560	372	58	editing	edit	VERB
bracis-33560	372	59	open	open	ADJ
bracis-33560	372	60	access	access	NOUN
bracis-33560	372	61	publishing	publishing	NOUN
bracis-33560	372	62	products	product	NOUN
bracis-33560	372	63	and	and	CCONJ
bracis-33560	372	64	services	service	NOUN
bracis-33560	372	65	our	our	PRON
bracis-33560	372	66	products	product	NOUN
bracis-33560	372	67	librarians	librarian	VERB
bracis-33560	372	68	societies	society	NOUN
bracis-33560	372	69	partners	partner	NOUN
bracis-33560	372	70	and	and	CCONJ
bracis-33560	372	71	advertisers	advertiser	NOUN
bracis-33560	372	72	our	our	PRON
bracis-33560	372	73	brands	brand	NOUN
bracis-33560	372	74	springer	springer	NOUN
bracis-33560	372	75	nature	nature	PROPN
bracis-33560	372	76	portfolio	portfolio	PROPN
bracis-33560	372	77	bmc	bmc	PROPN
bracis-33560	372	78	palgrave	palgrave	PROPN
bracis-33560	372	79	macmillan	macmillan	PROPN
bracis-33560	372	80	apress	apress	PROPN
bracis-33560	372	81	discover	discover	VERB
bracis-33560	372	82	your	your	PRON
bracis-33560	372	83	privacy	privacy	NOUN
bracis-33560	372	84	choices	choice	NOUN
bracis-33560	372	85	/	/	SYM
bracis-33560	372	86	manage	manage	NOUN
bracis-33560	372	87	cookies	cookie	NOUN
bracis-33560	372	88	your	your	PRON
bracis-33560	372	89	us	us	PROPN
bracis-33560	373	1	state	state	NOUN
bracis-33560	373	2	privacy	privacy	NOUN
bracis-33560	373	3	rights	right	NOUN
bracis-33560	373	4	accessibility	accessibility	NOUN
bracis-33560	373	5	statement	statement	NOUN
bracis-33560	373	6	terms	term	NOUN
bracis-33560	373	7	and	and	CCONJ
bracis-33560	373	8	conditions	condition	NOUN
bracis-33560	373	9	privacy	privacy	NOUN
bracis-33560	373	10	policy	policy	NOUN
bracis-33560	373	11	help	help	NOUN
bracis-33560	373	12	and	and	CCONJ
bracis-33560	373	13	support	support	VERB
bracis-33560	373	14	legal	legal	ADJ
bracis-33560	373	15	notice	notice	NOUN
bracis-33560	373	16	cancel	cancel	VERB
bracis-33560	373	17	contracts	contract	NOUN
bracis-33560	373	18	here	here	ADV
bracis-33560	373	19	129.74.145.123	129.74.145.123	NUM
bracis-33560	373	20	hesburgh	hesburgh	PROPN
bracis-33560	373	21	library	library	PROPN
bracis-33560	373	22	er	er	INTJ
bracis-33560	373	23	unit	unit	NOUN
bracis-33560	373	24	(	(	PUNCT
bracis-33560	373	25	3005732405	3005732405	NUM
bracis-33560	373	26	)	)	PUNCT
bracis-33560	373	27	northeast	northeast	ADJ
bracis-33560	373	28	research	research	NOUN
bracis-33560	373	29	libraries	library	NOUN
bracis-33560	373	30	(	(	PUNCT
bracis-33560	373	31	nerl	nerl	PROPN
bracis-33560	373	32	)	)	PUNCT
bracis-33560	373	33	(	(	PUNCT
bracis-33560	373	34	8200828607	8200828607	NUM
bracis-33560	373	35	)	)	PUNCT
bracis-33560	373	36	nerl	nerl	VERB
bracis-33560	373	37	ta	ta	X
bracis-33560	373	38	account	account	NOUN
bracis-33560	373	39	(	(	PUNCT
bracis-33560	373	40	3006206169	3006206169	NUM
bracis-33560	373	41	)	)	PUNCT
bracis-33560	373	42	university	university	NOUN
bracis-33560	373	43	of	of	ADP
bracis-33560	373	44	notre	notre	PROPN
bracis-33560	373	45	dame	dame	PROPN
bracis-33560	373	46	hesburgh	hesburgh	PROPN
bracis-33560	373	47	library	library	NOUN
bracis-33560	373	48	(	(	PUNCT
bracis-33560	373	49	3000184373	3000184373	NUM
bracis-33560	373	50	)	)	PUNCT
bracis-33560	374	1	©	©	ADP
bracis-33560	374	2	2025	2025	NUM
bracis-33560	374	3	springer	springer	NOUN
bracis-33560	374	4	nature	nature	NOUN
