id	sid	tid	token	lemma	pos
ejpam-1552	1	1	european	european	PROPN
ejpam-1552	1	2	journal	journal	PROPN
ejpam-1552	1	3	of	of	ADP
ejpam-1552	1	4	pure	pure	ADJ
ejpam-1552	1	5	and	and	CCONJ
ejpam-1552	1	6	applied	apply	VERB
ejpam-1552	1	7	mathematics	mathematic	NOUN
ejpam-1552	1	8	vol	vol	NOUN
ejpam-1552	1	9	.	.	PROPN
ejpam-1552	2	1	6	6	NUM
ejpam-1552	2	2	,	,	PUNCT
ejpam-1552	2	3	no	no	INTJ
ejpam-1552	2	4	.	.	NOUN
ejpam-1552	2	5	3	3	NUM
ejpam-1552	2	6	,	,	PUNCT
ejpam-1552	2	7	2013	2013	NUM
ejpam-1552	2	8	,	,	PUNCT
ejpam-1552	2	9	299	299	NUM
ejpam-1552	2	10	-	-	SYM
ejpam-1552	2	11	306	306	NUM
ejpam-1552	2	12	issn	issn	PROPN
ejpam-1552	2	13	1307	1307	NUM
ejpam-1552	2	14	-	-	SYM
ejpam-1552	2	15	5543	5543	NUM
ejpam-1552	2	16	–	–	PUNCT
ejpam-1552	2	17	www.ejpam.com	www.ejpam.com	X
ejpam-1552	2	18	decomposition	decomposition	NOUN
ejpam-1552	2	19	of	of	ADP
ejpam-1552	2	20	symmetry	symmetry	NOUN
ejpam-1552	2	21	model	model	NOUN
ejpam-1552	2	22	into	into	ADP
ejpam-1552	2	23	three	three	NUM
ejpam-1552	2	24	models	model	NOUN
ejpam-1552	2	25	for	for	ADP
ejpam-1552	2	26	cumulative	cumulative	ADJ
ejpam-1552	2	27	probabilities	probability	NOUN
ejpam-1552	2	28	in	in	ADP
ejpam-1552	2	29	square	square	ADJ
ejpam-1552	2	30	contingency	contingency	NOUN
ejpam-1552	2	31	tables	table	NOUN
ejpam-1552	2	32	kouji	kouji	PROPN
ejpam-1552	2	33	tahata1,∗	tahata1,∗	PROPN
ejpam-1552	2	34	,	,	PUNCT
ejpam-1552	2	35	kouji	kouji	PROPN
ejpam-1552	2	36	yamamoto2	yamamoto2	PROPN
ejpam-1552	2	37	,	,	PUNCT
ejpam-1552	2	38	sadao	sadao	VERB
ejpam-1552	2	39	tomizawa1	tomizawa1	NOUN
ejpam-1552	2	40	1	1	NUM
ejpam-1552	2	41	department	department	NOUN
ejpam-1552	2	42	of	of	ADP
ejpam-1552	2	43	information	information	NOUN
ejpam-1552	2	44	sciences	science	NOUN
ejpam-1552	2	45	,	,	PUNCT
ejpam-1552	2	46	faculty	faculty	NOUN
ejpam-1552	2	47	of	of	ADP
ejpam-1552	2	48	science	science	NOUN
ejpam-1552	2	49	and	and	CCONJ
ejpam-1552	2	50	technology	technology	NOUN
ejpam-1552	2	51	,	,	PUNCT
ejpam-1552	2	52	tokyo	tokyo	PROPN
ejpam-1552	2	53	university	university	PROPN
ejpam-1552	2	54	of	of	ADP
ejpam-1552	2	55	science	science	PROPN
ejpam-1552	2	56	,	,	PUNCT
ejpam-1552	2	57	noda	noda	PROPN
ejpam-1552	2	58	city	city	PROPN
ejpam-1552	2	59	,	,	PUNCT
ejpam-1552	2	60	chiba	chiba	PROPN
ejpam-1552	2	61	,	,	PUNCT
ejpam-1552	2	62	278	278	NUM
ejpam-1552	2	63	-	-	SYM
ejpam-1552	2	64	8510	8510	NUM
ejpam-1552	2	65	,	,	PUNCT
ejpam-1552	2	66	japan	japan	PROPN
ejpam-1552	2	67	2	2	NUM
ejpam-1552	2	68	department	department	NOUN
ejpam-1552	2	69	of	of	ADP
ejpam-1552	2	70	medical	medical	ADJ
ejpam-1552	2	71	innovation	innovation	NOUN
ejpam-1552	2	72	,	,	PUNCT
ejpam-1552	2	73	osaka	osaka	PROPN
ejpam-1552	2	74	university	university	NOUN
ejpam-1552	2	75	hospital	hospital	NOUN
ejpam-1552	2	76	,	,	PUNCT
ejpam-1552	2	77	2	2	NUM
ejpam-1552	2	78	-	-	SYM
ejpam-1552	2	79	15	15	NUM
ejpam-1552	2	80	,	,	PUNCT
ejpam-1552	2	81	yamadaoka	yamadaoka	NOUN
ejpam-1552	2	82	,	,	PUNCT
ejpam-1552	2	83	suita	suita	PROPN
ejpam-1552	2	84	,	,	PUNCT
ejpam-1552	2	85	osaka	osaka	PROPN
ejpam-1552	2	86	,	,	PUNCT
ejpam-1552	2	87	565	565	NUM
ejpam-1552	2	88	-	-	SYM
ejpam-1552	2	89	0871	0871	NUM
ejpam-1552	2	90	,	,	PUNCT
ejpam-1552	2	91	japan	japan	PROPN
ejpam-1552	2	92	abstract	abstract	PROPN
ejpam-1552	2	93	.	.	PUNCT
ejpam-1552	3	1	for	for	ADP
ejpam-1552	3	2	square	square	ADJ
ejpam-1552	3	3	contingency	contingency	NOUN
ejpam-1552	3	4	tables	table	NOUN
ejpam-1552	3	5	with	with	ADP
ejpam-1552	3	6	ordered	order	VERB
ejpam-1552	3	7	categories	category	NOUN
ejpam-1552	3	8	,	,	PUNCT
ejpam-1552	3	9	we	we	PRON
ejpam-1552	3	10	decompose	decompose	VERB
ejpam-1552	3	11	the	the	DET
ejpam-1552	3	12	symmetry	symmetry	NOUN
ejpam-1552	3	13	model	model	NOUN
ejpam-1552	3	14	into	into	ADP
ejpam-1552	3	15	three	three	NUM
ejpam-1552	3	16	models	model	NOUN
ejpam-1552	3	17	for	for	ADP
ejpam-1552	3	18	cumulative	cumulative	ADJ
ejpam-1552	3	19	probabilities	probability	NOUN
ejpam-1552	3	20	.	.	PUNCT
ejpam-1552	4	1	three	three	NUM
ejpam-1552	4	2	models	model	NOUN
ejpam-1552	4	3	are	be	AUX
ejpam-1552	4	4	the	the	DET
ejpam-1552	4	5	cumulative	cumulative	ADJ
ejpam-1552	4	6	two	two	NUM
ejpam-1552	4	7	ratios	ratio	NOUN
ejpam-1552	4	8	-	-	PUNCT
ejpam-1552	4	9	parameter	parameter	NOUN
ejpam-1552	4	10	symmetry	symmetry	NOUN
ejpam-1552	4	11	,	,	PUNCT
ejpam-1552	4	12	the	the	DET
ejpam-1552	4	13	global	global	ADJ
ejpam-1552	4	14	symmetry	symmetry	NOUN
ejpam-1552	4	15	,	,	PUNCT
ejpam-1552	4	16	and	and	CCONJ
ejpam-1552	4	17	the	the	DET
ejpam-1552	4	18	marginal	marginal	ADJ
ejpam-1552	4	19	means	mean	VERB
ejpam-1552	4	20	equality	equality	NOUN
ejpam-1552	4	21	models	model	NOUN
ejpam-1552	4	22	.	.	PUNCT
ejpam-1552	5	1	an	an	DET
ejpam-1552	5	2	example	example	NOUN
ejpam-1552	5	3	is	be	AUX
ejpam-1552	5	4	given	give	VERB
ejpam-1552	5	5	.	.	PUNCT
ejpam-1552	6	1	2010	2010	NUM
ejpam-1552	6	2	mathematics	mathematic	NOUN
ejpam-1552	6	3	subject	subject	NOUN
ejpam-1552	6	4	classifications	classification	NOUN
ejpam-1552	6	5	:	:	PUNCT
ejpam-1552	6	6	62h17	62h17	NUM
ejpam-1552	6	7	key	key	ADJ
ejpam-1552	6	8	words	word	NOUN
ejpam-1552	6	9	and	and	CCONJ
ejpam-1552	6	10	phrases	phrase	NOUN
ejpam-1552	6	11	:	:	PUNCT
ejpam-1552	6	12	decomposition	decomposition	NOUN
ejpam-1552	6	13	,	,	PUNCT
ejpam-1552	6	14	global	global	ADJ
ejpam-1552	6	15	symmetry	symmetry	NOUN
ejpam-1552	6	16	,	,	PUNCT
ejpam-1552	6	17	marginal	marginal	ADJ
ejpam-1552	6	18	mean	mean	NOUN
ejpam-1552	6	19	,	,	PUNCT
ejpam-1552	6	20	square	square	ADJ
ejpam-1552	6	21	contingency	contingency	NOUN
ejpam-1552	6	22	table	table	NOUN
ejpam-1552	6	23	,	,	PUNCT
ejpam-1552	6	24	symmetry	symmetry	NOUN
ejpam-1552	6	25	1	1	NUM
ejpam-1552	6	26	.	.	PUNCT
ejpam-1552	7	1	introduction	introduction	NOUN
ejpam-1552	7	2	for	for	ADP
ejpam-1552	7	3	an	an	DET
ejpam-1552	7	4	r	r	NOUN
ejpam-1552	7	5	×	×	NOUN
ejpam-1552	7	6	r	r	NOUN
ejpam-1552	7	7	square	square	ADJ
ejpam-1552	7	8	contingency	contingency	NOUN
ejpam-1552	7	9	table	table	NOUN
ejpam-1552	7	10	with	with	ADP
ejpam-1552	7	11	the	the	DET
ejpam-1552	7	12	same	same	ADJ
ejpam-1552	7	13	row	row	NOUN
ejpam-1552	7	14	and	and	CCONJ
ejpam-1552	7	15	column	column	NOUN
ejpam-1552	7	16	classifications	classification	NOUN
ejpam-1552	7	17	,	,	PUNCT
ejpam-1552	7	18	let	let	VERB
ejpam-1552	7	19	pi	pi	PROPN
ejpam-1552	7	20	j	j	PROPN
ejpam-1552	7	21	denote	denote	VERB
ejpam-1552	7	22	the	the	DET
ejpam-1552	7	23	probability	probability	NOUN
ejpam-1552	7	24	that	that	SCONJ
ejpam-1552	7	25	an	an	DET
ejpam-1552	7	26	observation	observation	NOUN
ejpam-1552	7	27	will	will	AUX
ejpam-1552	7	28	fall	fall	VERB
ejpam-1552	7	29	in	in	ADP
ejpam-1552	7	30	the	the	DET
ejpam-1552	7	31	ith	ith	NOUN
ejpam-1552	7	32	row	row	NOUN
ejpam-1552	7	33	and	and	CCONJ
ejpam-1552	7	34	jth	jth	PROPN
ejpam-1552	7	35	column	column	NOUN
ejpam-1552	7	36	of	of	ADP
ejpam-1552	7	37	the	the	DET
ejpam-1552	7	38	table	table	NOUN
ejpam-1552	7	39	(	(	PUNCT
ejpam-1552	7	40	i	i	NOUN
ejpam-1552	7	41	=	=	NOUN
ejpam-1552	7	42	1	1	NUM
ejpam-1552	7	43	,	,	PUNCT
ejpam-1552	7	44	.	.	PUNCT
ejpam-1552	7	45	.	.	PUNCT
ejpam-1552	8	1	.	.	PUNCT
ejpam-1552	9	1	,	,	PUNCT
ejpam-1552	9	2	r	r	NOUN
ejpam-1552	9	3	;	;	PUNCT
ejpam-1552	9	4	j	j	PROPN
ejpam-1552	9	5	=	=	SYM
ejpam-1552	9	6	1	1	NUM
ejpam-1552	9	7	,	,	PUNCT
ejpam-1552	9	8	.	.	PUNCT
ejpam-1552	9	9	.	.	PUNCT
ejpam-1552	9	10	.	.	PUNCT
ejpam-1552	10	1	,	,	PUNCT
ejpam-1552	10	2	r	r	NOUN
ejpam-1552	10	3	)	)	PUNCT
ejpam-1552	10	4	.	.	PUNCT
ejpam-1552	11	1	bowker	bowker	NOUN
ejpam-1552	12	1	[	[	X
ejpam-1552	12	2	3	3	X
ejpam-1552	12	3	]	]	PUNCT
ejpam-1552	12	4	considered	consider	VERB
ejpam-1552	12	5	the	the	DET
ejpam-1552	12	6	symmetry	symmetry	NOUN
ejpam-1552	12	7	(	(	PUNCT
ejpam-1552	12	8	s	s	NOUN
ejpam-1552	12	9	)	)	PUNCT
ejpam-1552	12	10	model	model	NOUN
ejpam-1552	12	11	defined	define	VERB
ejpam-1552	12	12	by	by	ADP
ejpam-1552	12	13	pi	pi	PROPN
ejpam-1552	12	14	j	j	PROPN
ejpam-1552	13	1	=	=	PUNCT
ejpam-1552	13	2	p	p	X
ejpam-1552	13	3	ji	ji	PROPN
ejpam-1552	13	4	(	(	PUNCT
ejpam-1552	13	5	i	i	PROPN
ejpam-1552	13	6	6=	6=	PROPN
ejpam-1552	13	7	j	j	PROPN
ejpam-1552	13	8	)	)	PUNCT
ejpam-1552	13	9	;	;	PUNCT
ejpam-1552	13	10	see	see	VERB
ejpam-1552	13	11	[	[	X
ejpam-1552	13	12	2	2	NUM
ejpam-1552	13	13	,	,	PUNCT
ejpam-1552	13	14	p.282	p.282	NOUN
ejpam-1552	13	15	]	]	PUNCT
ejpam-1552	13	16	.	.	PUNCT
ejpam-1552	14	1	caussinus	caussinus	NOUN
ejpam-1552	15	1	[	[	X
ejpam-1552	15	2	4	4	X
ejpam-1552	15	3	]	]	PUNCT
ejpam-1552	15	4	considered	consider	VERB
ejpam-1552	15	5	the	the	DET
ejpam-1552	15	6	quasi	quasi	NOUN
ejpam-1552	15	7	-	-	NOUN
ejpam-1552	15	8	symmetry	symmetry	NOUN
ejpam-1552	15	9	(	(	PUNCT
ejpam-1552	15	10	qs	qs	NOUN
ejpam-1552	15	11	)	)	PUNCT
ejpam-1552	15	12	model	model	NOUN
ejpam-1552	15	13	defined	define	VERB
ejpam-1552	15	14	by	by	ADP
ejpam-1552	15	15	pi	pi	PROPN
ejpam-1552	15	16	j	j	PROPN
ejpam-1552	16	1	=	=	PUNCT
ejpam-1552	16	2	µαiβ	µαiβ	PROPN
ejpam-1552	16	3	jψi	jψi	X
ejpam-1552	16	4	j	j	PROPN
ejpam-1552	16	5	(	(	PUNCT
ejpam-1552	16	6	i	i	NOUN
ejpam-1552	16	7	=	=	NOUN
ejpam-1552	16	8	1	1	NUM
ejpam-1552	16	9	,	,	PUNCT
ejpam-1552	16	10	.	.	PUNCT
ejpam-1552	16	11	.	.	PUNCT
ejpam-1552	16	12	.	.	PUNCT
ejpam-1552	17	1	,	,	PUNCT
ejpam-1552	17	2	r	r	NOUN
ejpam-1552	17	3	;	;	PUNCT
ejpam-1552	17	4	j	j	PROPN
ejpam-1552	17	5	=	=	SYM
ejpam-1552	17	6	1	1	NUM
ejpam-1552	17	7	,	,	PUNCT
ejpam-1552	17	8	.	.	PUNCT
ejpam-1552	17	9	.	.	PUNCT
ejpam-1552	17	10	.	.	PUNCT
ejpam-1552	18	1	,	,	PUNCT
ejpam-1552	18	2	r	r	X
ejpam-1552	18	3	)	)	PUNCT
ejpam-1552	18	4	,	,	PUNCT
ejpam-1552	18	5	where	where	SCONJ
ejpam-1552	18	6	ψi	ψi	ADP
ejpam-1552	18	7	j	j	PROPN
ejpam-1552	18	8	=	=	SYM
ejpam-1552	18	9	ψ	ψ	PROPN
ejpam-1552	18	10	ji	ji	PROPN
ejpam-1552	18	11	.	.	PUNCT
ejpam-1552	19	1	a	a	DET
ejpam-1552	19	2	special	special	ADJ
ejpam-1552	19	3	case	case	NOUN
ejpam-1552	19	4	of	of	ADP
ejpam-1552	19	5	qs	qs	PROPN
ejpam-1552	19	6	model	model	NOUN
ejpam-1552	19	7	obtained	obtain	VERB
ejpam-1552	19	8	by	by	ADP
ejpam-1552	19	9	putting	put	VERB
ejpam-1552	19	10	{	{	PUNCT
ejpam-1552	19	11	αi	αi	NOUN
ejpam-1552	19	12	=	=	PUNCT
ejpam-1552	19	13	βi	βi	X
ejpam-1552	19	14	}	}	PUNCT
ejpam-1552	19	15	is	be	AUX
ejpam-1552	19	16	the	the	DET
ejpam-1552	19	17	s	s	PROPN
ejpam-1552	19	18	model	model	NOUN
ejpam-1552	19	19	.	.	PUNCT
ejpam-1552	20	1	the	the	DET
ejpam-1552	20	2	marginal	marginal	ADJ
ejpam-1552	20	3	homogeneity	homogeneity	NOUN
ejpam-1552	20	4	(	(	PUNCT
ejpam-1552	20	5	mh	mh	NOUN
ejpam-1552	20	6	)	)	PUNCT
ejpam-1552	20	7	model	model	NOUN
ejpam-1552	20	8	is	be	AUX
ejpam-1552	20	9	defined	define	VERB
ejpam-1552	20	10	by	by	ADP
ejpam-1552	20	11	pi	pi	NOUN
ejpam-1552	20	12	·	·	PUNCT
ejpam-1552	20	13	=	=	PUNCT
ejpam-1552	20	14	p·i	p·i	X
ejpam-1552	20	15	(	(	PUNCT
ejpam-1552	20	16	i	i	NOUN
ejpam-1552	20	17	=	=	NOUN
ejpam-1552	20	18	1	1	NUM
ejpam-1552	20	19	,	,	PUNCT
ejpam-1552	20	20	.	.	PUNCT
ejpam-1552	20	21	.	.	PUNCT
ejpam-1552	20	22	.	.	PUNCT
ejpam-1552	20	23	,	,	PUNCT
ejpam-1552	20	24	r	r	NOUN
ejpam-1552	20	25	)	)	PUNCT
ejpam-1552	20	26	,	,	PUNCT
ejpam-1552	20	27	∗corresponding	∗corresponde	VERB
ejpam-1552	20	28	author	author	NOUN
ejpam-1552	20	29	.	.	PUNCT
ejpam-1552	21	1	email	email	NOUN
ejpam-1552	21	2	addresses	address	NOUN
ejpam-1552	21	3	:	:	PUNCT
ejpam-1552	21	4	kouji_tahata@is.noda.tus.ac.jp	kouji_tahata@is.noda.tus.ac.jp	PROPN
ejpam-1552	21	5	(	(	PUNCT
ejpam-1552	21	6	k.	k.	PROPN
ejpam-1552	21	7	tahata	tahata	PROPN
ejpam-1552	21	8	)	)	PUNCT
ejpam-1552	21	9	,	,	PUNCT
ejpam-1552	21	10	yamamoto-k@hp-crc.med.osaka-u.ac.jp	yamamoto-k@hp-crc.med.osaka-u.ac.jp	PROPN
ejpam-1552	21	11	(	(	PUNCT
ejpam-1552	21	12	k.	k.	PROPN
ejpam-1552	21	13	yamamoto	yamamoto	PROPN
ejpam-1552	21	14	)	)	PUNCT
ejpam-1552	21	15	,	,	PUNCT
ejpam-1552	21	16	tomizawa@is.noda.tus.ac.jp	tomizawa@is.noda.tus.ac.jp	PRON
ejpam-1552	21	17	(	(	PUNCT
ejpam-1552	21	18	s.	s.	PROPN
ejpam-1552	21	19	tomizawa	tomizawa	PROPN
ejpam-1552	21	20	)	)	PUNCT
ejpam-1552	21	21	http://www.ejpam.com	http://www.ejpam.com	X
ejpam-1552	22	1	299	299	NUM
ejpam-1552	22	2	c	c	X
ejpam-1552	22	3	©	©	PROPN
ejpam-1552	22	4	2013	2013	NUM
ejpam-1552	22	5	ejpam	ejpam	NOUN
ejpam-1552	22	6	all	all	DET
ejpam-1552	22	7	rights	right	NOUN
ejpam-1552	22	8	reserved	reserve	VERB
ejpam-1552	22	9	.	.	PUNCT
ejpam-1552	23	1	k.	k.	PROPN
ejpam-1552	23	2	tahata	tahata	PROPN
ejpam-1552	23	3	,	,	PUNCT
ejpam-1552	23	4	k.	k.	PROPN
ejpam-1552	23	5	yamamoto	yamamoto	PROPN
ejpam-1552	23	6	,	,	PUNCT
ejpam-1552	23	7	s.	s.	PROPN
ejpam-1552	23	8	tomizawa	tomizawa	PROPN
ejpam-1552	23	9	/	/	SYM
ejpam-1552	23	10	eur	eur	PROPN
ejpam-1552	23	11	.	.	PUNCT
ejpam-1552	24	1	j.	j.	PROPN
ejpam-1552	24	2	pure	pure	PROPN
ejpam-1552	24	3	appl	appl	PROPN
ejpam-1552	24	4	.	.	PROPN
ejpam-1552	24	5	math	math	PROPN
ejpam-1552	24	6	,	,	PUNCT
ejpam-1552	24	7	6	6	NUM
ejpam-1552	24	8	(	(	PUNCT
ejpam-1552	24	9	2013	2013	NUM
ejpam-1552	24	10	)	)	PUNCT
ejpam-1552	24	11	,	,	PUNCT
ejpam-1552	24	12	299	299	NUM
ejpam-1552	24	13	-	-	SYM
ejpam-1552	24	14	306	306	NUM
ejpam-1552	24	15	300	300	NUM
ejpam-1552	24	16	where	where	SCONJ
ejpam-1552	24	17	pi	pi	NOUN
ejpam-1552	24	18	·	·	PUNCT
ejpam-1552	25	1	=	=	PUNCT
ejpam-1552	25	2	r	r	NOUN
ejpam-1552	25	3	∑	∑	PUNCT
ejpam-1552	25	4	t=1	t=1	PROPN
ejpam-1552	25	5	pi	pi	PROPN
ejpam-1552	25	6	t	t	PROPN
ejpam-1552	25	7	,	,	PUNCT
ejpam-1552	25	8	p·i	p·i	PROPN
ejpam-1552	25	9	=	=	PUNCT
ejpam-1552	26	1	r	r	NOUN
ejpam-1552	26	2	∑	∑	PUNCT
ejpam-1552	26	3	s=1	s=1	X
ejpam-1552	26	4	psi	psi	NOUN
ejpam-1552	26	5	;	;	PUNCT
ejpam-1552	26	6	see	see	VERB
ejpam-1552	26	7	[	[	X
ejpam-1552	26	8	8	8	NUM
ejpam-1552	26	9	]	]	PUNCT
ejpam-1552	26	10	.	.	PUNCT
ejpam-1552	27	1	caussinus	caussinus	NOUN
ejpam-1552	28	1	[	[	X
ejpam-1552	28	2	4	4	X
ejpam-1552	28	3	]	]	PUNCT
ejpam-1552	28	4	gave	give	VERB
ejpam-1552	28	5	the	the	DET
ejpam-1552	28	6	following	following	NOUN
ejpam-1552	28	7	theorem	theorem	PROPN
ejpam-1552	28	8	.	.	PUNCT
ejpam-1552	28	9	theorem	theorem	NOUN
ejpam-1552	28	10	1	1	NUM
ejpam-1552	28	11	.	.	PUNCT
ejpam-1552	29	1	the	the	DET
ejpam-1552	29	2	s	s	PROPN
ejpam-1552	29	3	model	model	NOUN
ejpam-1552	29	4	holds	hold	VERB
ejpam-1552	29	5	if	if	SCONJ
ejpam-1552	29	6	and	and	CCONJ
ejpam-1552	29	7	only	only	ADV
ejpam-1552	29	8	if	if	SCONJ
ejpam-1552	29	9	both	both	PRON
ejpam-1552	29	10	the	the	DET
ejpam-1552	29	11	qs	qs	NOUN
ejpam-1552	29	12	and	and	CCONJ
ejpam-1552	29	13	mh	mh	PROPN
ejpam-1552	29	14	models	model	NOUN
ejpam-1552	29	15	hold	hold	VERB
ejpam-1552	29	16	.	.	PUNCT
ejpam-1552	30	1	tomizawa	tomizawa	PROPN
ejpam-1552	31	1	[	[	X
ejpam-1552	31	2	11	11	NUM
ejpam-1552	31	3	]	]	PUNCT
ejpam-1552	31	4	considered	consider	VERB
ejpam-1552	31	5	the	the	DET
ejpam-1552	31	6	two	two	NUM
ejpam-1552	31	7	ratios	ratio	NOUN
ejpam-1552	31	8	-	-	PUNCT
ejpam-1552	31	9	parameter	parameter	NOUN
ejpam-1552	31	10	symmetry	symmetry	NOUN
ejpam-1552	31	11	(	(	PUNCT
ejpam-1552	31	12	2rps	2rps	NOUN
ejpam-1552	31	13	)	)	PUNCT
ejpam-1552	31	14	model	model	NOUN
ejpam-1552	31	15	defined	define	VERB
ejpam-1552	31	16	by	by	ADP
ejpam-1552	31	17	pi	pi	PROPN
ejpam-1552	31	18	j	j	PROPN
ejpam-1552	31	19	p	p	X
ejpam-1552	31	20	ji	ji	PROPN
ejpam-1552	31	21	=	=	PUNCT
ejpam-1552	31	22	γφ	γφ	PROPN
ejpam-1552	31	23	j−i	j−i	PROPN
ejpam-1552	31	24	(	(	PUNCT
ejpam-1552	31	25	i	i	PRON
ejpam-1552	31	26	<	<	X
ejpam-1552	31	27	j	j	PROPN
ejpam-1552	31	28	)	)	PUNCT
ejpam-1552	31	29	.	.	PUNCT
ejpam-1552	32	1	special	special	ADJ
ejpam-1552	32	2	cases	case	NOUN
ejpam-1552	32	3	of	of	ADP
ejpam-1552	32	4	2rps	2rps	NUM
ejpam-1552	32	5	model	model	NOUN
ejpam-1552	32	6	obtained	obtain	VERB
ejpam-1552	32	7	by	by	ADP
ejpam-1552	32	8	putting	put	VERB
ejpam-1552	32	9	φ	φ	PROPN
ejpam-1552	32	10	=	=	SYM
ejpam-1552	32	11	1	1	NUM
ejpam-1552	32	12	and	and	CCONJ
ejpam-1552	32	13	γ	γ	X
ejpam-1552	32	14	=	=	SYM
ejpam-1552	32	15	1	1	NUM
ejpam-1552	32	16	are	be	AUX
ejpam-1552	32	17	mccullagh	mccullagh	NOUN
ejpam-1552	32	18	’s	’s	PART
ejpam-1552	32	19	[	[	X
ejpam-1552	32	20	6	6	NUM
ejpam-1552	32	21	]	]	X
ejpam-1552	32	22	conditional	conditional	ADJ
ejpam-1552	32	23	symmetry	symmetry	NOUN
ejpam-1552	32	24	(	(	PUNCT
ejpam-1552	32	25	cs	cs	ADJ
ejpam-1552	32	26	)	)	PUNCT
ejpam-1552	32	27	and	and	CCONJ
ejpam-1552	32	28	agresti	agresti	PROPN
ejpam-1552	32	29	’s	’s	PART
ejpam-1552	32	30	[	[	X
ejpam-1552	32	31	1	1	NUM
ejpam-1552	32	32	]	]	PUNCT
ejpam-1552	32	33	linear	linear	ADJ
ejpam-1552	32	34	diagonals	diagonal	NOUN
ejpam-1552	32	35	-	-	PUNCT
ejpam-1552	32	36	parameter	parameter	NOUN
ejpam-1552	32	37	symmetry	symmetry	NOUN
ejpam-1552	32	38	(	(	PUNCT
ejpam-1552	32	39	ldps	ldps	PROPN
ejpam-1552	32	40	)	)	PUNCT
ejpam-1552	32	41	models	model	NOUN
ejpam-1552	32	42	,	,	PUNCT
ejpam-1552	32	43	respectively	respectively	ADV
ejpam-1552	32	44	.	.	PUNCT
ejpam-1552	33	1	define	define	VERB
ejpam-1552	33	2	the	the	DET
ejpam-1552	33	3	global	global	ADJ
ejpam-1552	33	4	symmetry	symmetry	NOUN
ejpam-1552	33	5	(	(	PUNCT
ejpam-1552	33	6	gs	gs	NOUN
ejpam-1552	33	7	)	)	PUNCT
ejpam-1552	33	8	model	model	NOUN
ejpam-1552	33	9	by	by	ADP
ejpam-1552	33	10	δu	δu	NOUN
ejpam-1552	33	11	=	=	SYM
ejpam-1552	33	12	δl	δl	PROPN
ejpam-1552	33	13	,	,	PUNCT
ejpam-1552	33	14	where	where	SCONJ
ejpam-1552	33	15	δu	δu	ADP
ejpam-1552	33	16	=	=	SYM
ejpam-1552	33	17	∑∑	∑∑	PROPN
ejpam-1552	33	18	i	i	PRON
ejpam-1552	33	19	<	<	X
ejpam-1552	33	20	j	j	PROPN
ejpam-1552	33	21	pi	pi	PROPN
ejpam-1552	33	22	j	j	PROPN
ejpam-1552	33	23	,	,	PUNCT
ejpam-1552	33	24	δl	δl	PROPN
ejpam-1552	33	25	=	=	SYM
ejpam-1552	33	26	∑∑	∑∑	PROPN
ejpam-1552	33	27	i	i	PRON
ejpam-1552	33	28	>	>	X
ejpam-1552	33	29	j	j	PROPN
ejpam-1552	33	30	pi	pi	PROPN
ejpam-1552	33	31	j	j	PROPN
ejpam-1552	33	32	.	.	PUNCT
ejpam-1552	34	1	let	let	VERB
ejpam-1552	34	2	x	x	PRON
ejpam-1552	34	3	and	and	CCONJ
ejpam-1552	34	4	y	y	PROPN
ejpam-1552	34	5	denote	denote	VERB
ejpam-1552	34	6	the	the	DET
ejpam-1552	34	7	row	row	NOUN
ejpam-1552	34	8	and	and	CCONJ
ejpam-1552	34	9	column	column	NOUN
ejpam-1552	34	10	variables	variable	NOUN
ejpam-1552	34	11	,	,	PUNCT
ejpam-1552	34	12	respectively	respectively	ADV
ejpam-1552	34	13	.	.	PUNCT
ejpam-1552	35	1	define	define	VERB
ejpam-1552	35	2	the	the	DET
ejpam-1552	35	3	marginal	marginal	ADJ
ejpam-1552	35	4	means	mean	VERB
ejpam-1552	35	5	equality	equality	NOUN
ejpam-1552	35	6	(	(	PUNCT
ejpam-1552	35	7	me	me	NOUN
ejpam-1552	35	8	)	)	PUNCT
ejpam-1552	35	9	model	model	NOUN
ejpam-1552	35	10	by	by	ADP
ejpam-1552	35	11	e(x	e(x	NOUN
ejpam-1552	35	12	)	)	PUNCT
ejpam-1552	36	1	=	=	PUNCT
ejpam-1552	36	2	e(y	e(y	ADJ
ejpam-1552	36	3	)	)	PUNCT
ejpam-1552	36	4	,	,	PUNCT
ejpam-1552	36	5	where	where	SCONJ
ejpam-1552	36	6	e(x	e(x	NOUN
ejpam-1552	36	7	)	)	PUNCT
ejpam-1552	37	1	=	=	PUNCT
ejpam-1552	37	2	r	r	NOUN
ejpam-1552	37	3	∑	∑	PROPN
ejpam-1552	37	4	i=1	i=1	PROPN
ejpam-1552	37	5	ipi	ipi	PROPN
ejpam-1552	37	6	·	·	PROPN
ejpam-1552	37	7	,	,	PUNCT
ejpam-1552	37	8	e(y	e(y	ADJ
ejpam-1552	37	9	)	)	PUNCT
ejpam-1552	38	1	=	=	PUNCT
ejpam-1552	38	2	r	r	NOUN
ejpam-1552	38	3	∑	∑	PUNCT
ejpam-1552	38	4	i=1	i=1	PROPN
ejpam-1552	38	5	ip·i	ip·i	PROPN
ejpam-1552	38	6	.	.	PUNCT
ejpam-1552	39	1	yamamoto	yamamoto	PROPN
ejpam-1552	39	2	,	,	PUNCT
ejpam-1552	39	3	iwashita	iwashita	NOUN
ejpam-1552	39	4	and	and	CCONJ
ejpam-1552	39	5	tomizawa	tomizawa	PRON
ejpam-1552	40	1	[	[	X
ejpam-1552	40	2	15	15	NUM
ejpam-1552	40	3	]	]	PUNCT
ejpam-1552	40	4	,	,	PUNCT
ejpam-1552	40	5	and	and	CCONJ
ejpam-1552	40	6	tahata	tahata	PROPN
ejpam-1552	40	7	,	,	PUNCT
ejpam-1552	40	8	yamamoto	yamamoto	NOUN
ejpam-1552	40	9	and	and	CCONJ
ejpam-1552	40	10	tomizawa	tomizawa	PROPN
ejpam-1552	41	1	[	[	X
ejpam-1552	41	2	10	10	NUM
ejpam-1552	41	3	]	]	PUNCT
ejpam-1552	41	4	gave	give	VERB
ejpam-1552	41	5	the	the	DET
ejpam-1552	41	6	following	follow	VERB
ejpam-1552	41	7	theorem	theorem	PROPN
ejpam-1552	41	8	.	.	PUNCT
ejpam-1552	42	1	theorem	theorem	NOUN
ejpam-1552	42	2	2	2	NUM
ejpam-1552	42	3	.	.	PUNCT
ejpam-1552	43	1	the	the	DET
ejpam-1552	43	2	s	s	PROPN
ejpam-1552	43	3	model	model	NOUN
ejpam-1552	43	4	holds	hold	VERB
ejpam-1552	43	5	if	if	SCONJ
ejpam-1552	43	6	and	and	CCONJ
ejpam-1552	43	7	only	only	ADV
ejpam-1552	43	8	if	if	SCONJ
ejpam-1552	43	9	both	both	CCONJ
ejpam-1552	43	10	the	the	DET
ejpam-1552	43	11	ldps	ldps	NOUN
ejpam-1552	43	12	and	and	CCONJ
ejpam-1552	43	13	me	i	PRON
ejpam-1552	43	14	models	model	NOUN
ejpam-1552	43	15	hold	hold	VERB
ejpam-1552	43	16	.	.	PUNCT
ejpam-1552	44	1	tahata	tahata	PROPN
ejpam-1552	44	2	and	and	CCONJ
ejpam-1552	44	3	tomizawa	tomizawa	NOUN
ejpam-1552	45	1	[	[	X
ejpam-1552	45	2	9	9	NUM
ejpam-1552	45	3	]	]	PUNCT
ejpam-1552	45	4	gave	give	VERB
ejpam-1552	45	5	the	the	DET
ejpam-1552	45	6	theorem	theorem	NOUN
ejpam-1552	45	7	as	as	SCONJ
ejpam-1552	45	8	follows	follow	VERB
ejpam-1552	45	9	.	.	PUNCT
ejpam-1552	46	1	theorem	theorem	NOUN
ejpam-1552	46	2	3	3	NUM
ejpam-1552	46	3	.	.	PUNCT
ejpam-1552	47	1	the	the	DET
ejpam-1552	47	2	s	s	PROPN
ejpam-1552	47	3	model	model	NOUN
ejpam-1552	47	4	holds	hold	VERB
ejpam-1552	47	5	if	if	SCONJ
ejpam-1552	47	6	and	and	CCONJ
ejpam-1552	47	7	only	only	ADV
ejpam-1552	47	8	if	if	SCONJ
ejpam-1552	47	9	all	all	DET
ejpam-1552	47	10	the	the	DET
ejpam-1552	47	11	2rps	2rp	NOUN
ejpam-1552	47	12	,	,	PUNCT
ejpam-1552	47	13	gs	gs	PROPN
ejpam-1552	47	14	and	and	CCONJ
ejpam-1552	47	15	me	i	PRON
ejpam-1552	47	16	models	model	NOUN
ejpam-1552	47	17	hold	hold	VERB
ejpam-1552	47	18	.	.	PUNCT
ejpam-1552	48	1	let	let	VERB
ejpam-1552	48	2	gi	gi	INTJ
ejpam-1552	48	3	j	j	X
ejpam-1552	49	1	=	=	PUNCT
ejpam-1552	49	2	i	i	INTJ
ejpam-1552	49	3	∑	∑	PUNCT
ejpam-1552	49	4	s=1	s=1	PUNCT
ejpam-1552	49	5	r	r	NOUN
ejpam-1552	49	6	∑	∑	PUNCT
ejpam-1552	49	7	t=	t=	ADJ
ejpam-1552	49	8	j	j	PROPN
ejpam-1552	49	9	pst	pst	NOUN
ejpam-1552	49	10	(	(	PUNCT
ejpam-1552	49	11	i	i	PRON
ejpam-1552	49	12	<	<	X
ejpam-1552	49	13	j	j	PROPN
ejpam-1552	49	14	)	)	PUNCT
ejpam-1552	49	15	,	,	PUNCT
ejpam-1552	49	16	and	and	CCONJ
ejpam-1552	50	1	gi	gi	X
ejpam-1552	50	2	j	j	NOUN
ejpam-1552	51	1	=	=	SYM
ejpam-1552	51	2	r	r	X
ejpam-1552	51	3	∑	∑	PUNCT
ejpam-1552	51	4	s	s	PROPN
ejpam-1552	51	5	=	=	PROPN
ejpam-1552	51	6	i	i	NOUN
ejpam-1552	51	7	j	j	NOUN
ejpam-1552	51	8	∑	∑	PUNCT
ejpam-1552	51	9	t=1	t=1	PUNCT
ejpam-1552	51	10	pst	pst	VERB
ejpam-1552	51	11	(	(	PUNCT
ejpam-1552	51	12	i	i	PROPN
ejpam-1552	51	13	>	>	X
ejpam-1552	51	14	j	j	PROPN
ejpam-1552	51	15	)	)	PUNCT
ejpam-1552	51	16	.	.	PUNCT
ejpam-1552	52	1	k.	k.	PROPN
ejpam-1552	52	2	tahata	tahata	PROPN
ejpam-1552	52	3	,	,	PUNCT
ejpam-1552	52	4	k.	k.	PROPN
ejpam-1552	52	5	yamamoto	yamamoto	PROPN
ejpam-1552	52	6	,	,	PUNCT
ejpam-1552	52	7	s.	s.	PROPN
ejpam-1552	52	8	tomizawa	tomizawa	PROPN
ejpam-1552	52	9	/	/	SYM
ejpam-1552	52	10	eur	eur	PROPN
ejpam-1552	52	11	.	.	PUNCT
ejpam-1552	53	1	j.	j.	PROPN
ejpam-1552	53	2	pure	pure	PROPN
ejpam-1552	53	3	appl	appl	PROPN
ejpam-1552	53	4	.	.	PROPN
ejpam-1552	53	5	math	math	PROPN
ejpam-1552	53	6	,	,	PUNCT
ejpam-1552	53	7	6	6	NUM
ejpam-1552	53	8	(	(	PUNCT
ejpam-1552	53	9	2013	2013	NUM
ejpam-1552	53	10	)	)	PUNCT
ejpam-1552	53	11	,	,	PUNCT
ejpam-1552	53	12	299	299	NUM
ejpam-1552	53	13	-	-	SYM
ejpam-1552	53	14	306	306	NUM
ejpam-1552	53	15	301	301	NUM
ejpam-1552	53	16	note	note	NOUN
ejpam-1552	53	17	that	that	SCONJ
ejpam-1552	53	18	the	the	DET
ejpam-1552	53	19	s	s	NOUN
ejpam-1552	53	20	model	model	NOUN
ejpam-1552	53	21	may	may	AUX
ejpam-1552	53	22	be	be	AUX
ejpam-1552	53	23	expressed	express	VERB
ejpam-1552	53	24	as	as	ADP
ejpam-1552	53	25	gi	gi	VERB
ejpam-1552	53	26	j	j	PROPN
ejpam-1552	53	27	=	=	SYM
ejpam-1552	54	1	g	g	PROPN
ejpam-1552	54	2	ji	ji	PROPN
ejpam-1552	54	3	(	(	PUNCT
ejpam-1552	54	4	i	i	PROPN
ejpam-1552	54	5	6=	6=	PROPN
ejpam-1552	54	6	j	j	PROPN
ejpam-1552	54	7	)	)	PUNCT
ejpam-1552	54	8	.	.	PUNCT
ejpam-1552	55	1	the	the	DET
ejpam-1552	55	2	mh	mh	PROPN
ejpam-1552	55	3	model	model	NOUN
ejpam-1552	55	4	may	may	AUX
ejpam-1552	55	5	be	be	AUX
ejpam-1552	55	6	expressed	express	VERB
ejpam-1552	55	7	as	as	ADP
ejpam-1552	55	8	gi	gi	NOUN
ejpam-1552	55	9	,	,	PUNCT
ejpam-1552	55	10	i+1	i+1	NOUN
ejpam-1552	55	11	=	=	X
ejpam-1552	55	12	gi+1,i	gi+1,i	NOUN
ejpam-1552	55	13	(	(	PUNCT
ejpam-1552	55	14	i	i	NOUN
ejpam-1552	55	15	=	=	NOUN
ejpam-1552	55	16	1	1	NUM
ejpam-1552	55	17	,	,	PUNCT
ejpam-1552	55	18	.	.	PUNCT
ejpam-1552	55	19	.	.	PUNCT
ejpam-1552	56	1	.	.	PUNCT
ejpam-1552	57	1	,	,	PUNCT
ejpam-1552	57	2	r	r	NOUN
ejpam-1552	57	3	−	−	PROPN
ejpam-1552	57	4	1	1	NUM
ejpam-1552	57	5	)	)	PUNCT
ejpam-1552	57	6	.	.	PUNCT
ejpam-1552	58	1	miyamoto	miyamoto	NOUN
ejpam-1552	58	2	,	,	PUNCT
ejpam-1552	58	3	ohtsuka	ohtsuka	NOUN
ejpam-1552	58	4	and	and	CCONJ
ejpam-1552	58	5	tomizawa	tomizawa	PROPN
ejpam-1552	59	1	[	[	X
ejpam-1552	59	2	7	7	X
ejpam-1552	59	3	]	]	PUNCT
ejpam-1552	59	4	considered	consider	VERB
ejpam-1552	59	5	the	the	DET
ejpam-1552	59	6	cumulative	cumulative	ADJ
ejpam-1552	59	7	quasi	quasi	NOUN
ejpam-1552	59	8	-	-	NOUN
ejpam-1552	59	9	symmetry	symmetry	ADJ
ejpam-1552	59	10	(	(	PUNCT
ejpam-1552	59	11	cqs	cqs	NOUN
ejpam-1552	59	12	)	)	PUNCT
ejpam-1552	59	13	model	model	NOUN
ejpam-1552	59	14	defined	define	VERB
ejpam-1552	59	15	by	by	ADP
ejpam-1552	59	16	gi	gi	NUM
ejpam-1552	59	17	j	j	PROPN
ejpam-1552	60	1	=	=	PRON
ejpam-1552	60	2	µξiη	µξiη	PROPN
ejpam-1552	60	3	jψi	jψi	PROPN
ejpam-1552	61	1	j	j	PROPN
ejpam-1552	61	2	(	(	PUNCT
ejpam-1552	61	3	i	i	PROPN
ejpam-1552	61	4	6=	6=	PROPN
ejpam-1552	61	5	j	j	PROPN
ejpam-1552	61	6	)	)	PUNCT
ejpam-1552	61	7	,	,	PUNCT
ejpam-1552	61	8	where	where	SCONJ
ejpam-1552	61	9	ψi	ψi	ADP
ejpam-1552	61	10	j	j	PROPN
ejpam-1552	61	11	=	=	PROPN
ejpam-1552	61	12	ψ	ψ	X
ejpam-1552	61	13	ji	ji	PROPN
ejpam-1552	61	14	.	.	PUNCT
ejpam-1552	62	1	yamamoto	yamamoto	PROPN
ejpam-1552	62	2	,	,	PUNCT
ejpam-1552	62	3	ando	ando	PROPN
ejpam-1552	62	4	and	and	CCONJ
ejpam-1552	62	5	tomizawa	tomizawa	PROPN
ejpam-1552	63	1	[	[	X
ejpam-1552	63	2	16	16	NUM
ejpam-1552	63	3	]	]	PUNCT
ejpam-1552	63	4	gave	give	VERB
ejpam-1552	63	5	the	the	DET
ejpam-1552	63	6	following	following	NOUN
ejpam-1552	63	7	theorem	theorem	PROPN
ejpam-1552	63	8	.	.	PUNCT
ejpam-1552	63	9	theorem	theorem	NOUN
ejpam-1552	63	10	4	4	NUM
ejpam-1552	63	11	.	.	PUNCT
ejpam-1552	64	1	the	the	DET
ejpam-1552	64	2	s	s	PROPN
ejpam-1552	64	3	model	model	NOUN
ejpam-1552	64	4	holds	hold	VERB
ejpam-1552	64	5	if	if	SCONJ
ejpam-1552	64	6	and	and	CCONJ
ejpam-1552	64	7	only	only	ADV
ejpam-1552	64	8	if	if	SCONJ
ejpam-1552	64	9	both	both	CCONJ
ejpam-1552	64	10	the	the	DET
ejpam-1552	64	11	cqs	cqs	NOUN
ejpam-1552	64	12	and	and	CCONJ
ejpam-1552	64	13	mh	mh	PROPN
ejpam-1552	64	14	models	model	NOUN
ejpam-1552	64	15	hold	hold	VERB
ejpam-1552	64	16	.	.	PUNCT
ejpam-1552	65	1	miyamoto	miyamoto	NOUN
ejpam-1552	65	2	et	et	PROPN
ejpam-1552	65	3	al	al	PROPN
ejpam-1552	65	4	.	.	PUNCT
ejpam-1552	66	1	[	[	X
ejpam-1552	66	2	7	7	X
ejpam-1552	66	3	]	]	PUNCT
ejpam-1552	66	4	also	also	ADV
ejpam-1552	66	5	considered	consider	VERB
ejpam-1552	66	6	the	the	DET
ejpam-1552	66	7	cumulative	cumulative	ADJ
ejpam-1552	66	8	linear	linear	ADJ
ejpam-1552	66	9	diagonals	diagonal	NOUN
ejpam-1552	66	10	-	-	PUNCT
ejpam-1552	66	11	parameter	parameter	NOUN
ejpam-1552	66	12	symmetry	symmetry	NOUN
ejpam-1552	66	13	(	(	PUNCT
ejpam-1552	66	14	cldps	cldps	NOUN
ejpam-1552	66	15	)	)	PUNCT
ejpam-1552	66	16	model	model	NOUN
ejpam-1552	66	17	defined	define	VERB
ejpam-1552	66	18	by	by	ADP
ejpam-1552	66	19	gi	gi	NUM
ejpam-1552	66	20	j	j	PROPN
ejpam-1552	66	21	g	g	PROPN
ejpam-1552	66	22	ji	ji	PROPN
ejpam-1552	66	23	=	=	PROPN
ejpam-1552	66	24	θ	θ	PROPN
ejpam-1552	66	25	j−i	j−i	PROPN
ejpam-1552	66	26	(	(	PUNCT
ejpam-1552	66	27	i	i	PRON
ejpam-1552	66	28	<	<	X
ejpam-1552	66	29	j	j	PROPN
ejpam-1552	66	30	)	)	PUNCT
ejpam-1552	66	31	.	.	PUNCT
ejpam-1552	67	1	yamamoto	yamamoto	PROPN
ejpam-1552	67	2	and	and	CCONJ
ejpam-1552	67	3	tomizawa	tomizawa	PROPN
ejpam-1552	68	1	[	[	X
ejpam-1552	68	2	17	17	NUM
ejpam-1552	68	3	]	]	PUNCT
ejpam-1552	68	4	gave	give	VERB
ejpam-1552	68	5	the	the	DET
ejpam-1552	68	6	theorem	theorem	NOUN
ejpam-1552	68	7	as	as	SCONJ
ejpam-1552	68	8	follows	follow	VERB
ejpam-1552	68	9	.	.	PUNCT
ejpam-1552	69	1	theorem	theorem	ADJ
ejpam-1552	69	2	5	5	NUM
ejpam-1552	69	3	.	.	PUNCT
ejpam-1552	70	1	the	the	DET
ejpam-1552	70	2	s	s	PROPN
ejpam-1552	70	3	model	model	NOUN
ejpam-1552	70	4	holds	hold	VERB
ejpam-1552	70	5	if	if	SCONJ
ejpam-1552	70	6	and	and	CCONJ
ejpam-1552	70	7	only	only	ADV
ejpam-1552	70	8	if	if	SCONJ
ejpam-1552	70	9	both	both	PRON
ejpam-1552	70	10	the	the	DET
ejpam-1552	70	11	cldps	cldps	NOUN
ejpam-1552	70	12	and	and	CCONJ
ejpam-1552	70	13	me	i	PRON
ejpam-1552	70	14	models	model	NOUN
ejpam-1552	70	15	hold	hold	VERB
ejpam-1552	70	16	.	.	PUNCT
ejpam-1552	71	1	tomizawa	tomizawa	PROPN
ejpam-1552	71	2	,	,	PUNCT
ejpam-1552	71	3	miyamoto	miyamoto	NOUN
ejpam-1552	71	4	,	,	PUNCT
ejpam-1552	71	5	yamamoto	yamamoto	NOUN
ejpam-1552	71	6	and	and	CCONJ
ejpam-1552	71	7	sugiyama	sugiyama	NOUN
ejpam-1552	71	8	[	[	X
ejpam-1552	71	9	13	13	NUM
ejpam-1552	71	10	]	]	PUNCT
ejpam-1552	71	11	considered	consider	VERB
ejpam-1552	71	12	the	the	DET
ejpam-1552	71	13	cumulative	cumulative	ADJ
ejpam-1552	71	14	two	two	NUM
ejpam-1552	71	15	ratiosparameter	ratiosparameter	NOUN
ejpam-1552	71	16	symmetry	symmetry	NOUN
ejpam-1552	71	17	(	(	PUNCT
ejpam-1552	71	18	c2rps	c2rps	PROPN
ejpam-1552	71	19	)	)	PUNCT
ejpam-1552	71	20	model	model	NOUN
ejpam-1552	71	21	defined	define	VERB
ejpam-1552	71	22	by	by	ADP
ejpam-1552	71	23	gi	gi	NUM
ejpam-1552	71	24	j	j	PROPN
ejpam-1552	71	25	g	g	PROPN
ejpam-1552	71	26	ji	ji	PROPN
ejpam-1552	72	1	=	=	PUNCT
ejpam-1552	72	2	γθ	γθ	PROPN
ejpam-1552	72	3	j−i	j−i	PROPN
ejpam-1552	72	4	(	(	PUNCT
ejpam-1552	72	5	i	i	PRON
ejpam-1552	72	6	<	<	X
ejpam-1552	72	7	j	j	PROPN
ejpam-1552	72	8	)	)	PUNCT
ejpam-1552	72	9	.	.	PUNCT
ejpam-1552	73	1	we	we	PRON
ejpam-1552	73	2	are	be	AUX
ejpam-1552	73	3	now	now	ADV
ejpam-1552	73	4	interested	interested	ADJ
ejpam-1552	73	5	in	in	ADP
ejpam-1552	73	6	whether	whether	SCONJ
ejpam-1552	73	7	or	or	CCONJ
ejpam-1552	73	8	not	not	PART
ejpam-1552	73	9	theorem	theorem	VERB
ejpam-1552	73	10	3	3	NUM
ejpam-1552	73	11	with	with	SCONJ
ejpam-1552	73	12	the	the	DET
ejpam-1552	73	13	2rps	2rps	NUM
ejpam-1552	73	14	model	model	NOUN
ejpam-1552	73	15	replaced	replace	VERB
ejpam-1552	73	16	by	by	ADP
ejpam-1552	73	17	the	the	DET
ejpam-1552	73	18	c2rps	c2rps	PROPN
ejpam-1552	73	19	model	model	NOUN
ejpam-1552	73	20	holds	hold	VERB
ejpam-1552	73	21	.	.	PUNCT
ejpam-1552	74	1	the	the	DET
ejpam-1552	74	2	purpose	purpose	NOUN
ejpam-1552	74	3	of	of	ADP
ejpam-1552	74	4	this	this	DET
ejpam-1552	74	5	paper	paper	NOUN
ejpam-1552	74	6	is	be	AUX
ejpam-1552	74	7	to	to	PART
ejpam-1552	74	8	decompose	decompose	VERB
ejpam-1552	74	9	the	the	DET
ejpam-1552	74	10	s	s	NOUN
ejpam-1552	74	11	model	model	NOUN
ejpam-1552	74	12	into	into	ADP
ejpam-1552	74	13	three	three	NUM
ejpam-1552	74	14	models	model	NOUN
ejpam-1552	74	15	,	,	PUNCT
ejpam-1552	74	16	i.e.	i.e.	X
ejpam-1552	74	17	,	,	PUNCT
ejpam-1552	74	18	the	the	DET
ejpam-1552	74	19	c2rps	c2rps	PROPN
ejpam-1552	74	20	,	,	PUNCT
ejpam-1552	74	21	the	the	DET
ejpam-1552	74	22	gs	gs	NOUN
ejpam-1552	74	23	,	,	PUNCT
ejpam-1552	74	24	and	and	CCONJ
ejpam-1552	74	25	the	the	DET
ejpam-1552	74	26	me	me	PROPN
ejpam-1552	74	27	models	model	NOUN
ejpam-1552	74	28	.	.	PUNCT
ejpam-1552	75	1	2	2	X
ejpam-1552	75	2	.	.	X
ejpam-1552	75	3	new	new	ADJ
ejpam-1552	75	4	decomposition	decomposition	NOUN
ejpam-1552	75	5	of	of	ADP
ejpam-1552	75	6	symmetry	symmetry	NOUN
ejpam-1552	75	7	we	we	PRON
ejpam-1552	75	8	can	can	AUX
ejpam-1552	75	9	obtain	obtain	VERB
ejpam-1552	75	10	a	a	DET
ejpam-1552	75	11	new	new	ADJ
ejpam-1552	75	12	decomposition	decomposition	NOUN
ejpam-1552	75	13	of	of	ADP
ejpam-1552	75	14	the	the	DET
ejpam-1552	75	15	symmetry	symmetry	NOUN
ejpam-1552	75	16	model	model	NOUN
ejpam-1552	75	17	as	as	SCONJ
ejpam-1552	75	18	follows	follow	VERB
ejpam-1552	75	19	.	.	PUNCT
ejpam-1552	76	1	theorem	theorem	ADJ
ejpam-1552	76	2	6	6	NUM
ejpam-1552	76	3	.	.	PUNCT
ejpam-1552	77	1	the	the	DET
ejpam-1552	77	2	s	s	PROPN
ejpam-1552	77	3	model	model	NOUN
ejpam-1552	77	4	holds	hold	VERB
ejpam-1552	77	5	if	if	SCONJ
ejpam-1552	77	6	and	and	CCONJ
ejpam-1552	77	7	only	only	ADV
ejpam-1552	77	8	if	if	SCONJ
ejpam-1552	77	9	all	all	DET
ejpam-1552	77	10	the	the	DET
ejpam-1552	77	11	c2rps	c2rps	PROPN
ejpam-1552	77	12	,	,	PUNCT
ejpam-1552	77	13	gs	gs	NOUN
ejpam-1552	77	14	,	,	PUNCT
ejpam-1552	77	15	and	and	CCONJ
ejpam-1552	77	16	me	i	PRON
ejpam-1552	77	17	models	model	NOUN
ejpam-1552	77	18	hold	hold	VERB
ejpam-1552	77	19	.	.	PUNCT
ejpam-1552	78	1	proof	proof	NOUN
ejpam-1552	78	2	.	.	PUNCT
ejpam-1552	79	1	if	if	SCONJ
ejpam-1552	79	2	the	the	DET
ejpam-1552	79	3	s	s	PROPN
ejpam-1552	79	4	model	model	NOUN
ejpam-1552	79	5	holds	hold	VERB
ejpam-1552	79	6	,	,	PUNCT
ejpam-1552	79	7	then	then	ADV
ejpam-1552	79	8	all	all	DET
ejpam-1552	79	9	the	the	DET
ejpam-1552	79	10	c2rps	c2rps	PROPN
ejpam-1552	79	11	,	,	PUNCT
ejpam-1552	79	12	gs	gs	NOUN
ejpam-1552	79	13	,	,	PUNCT
ejpam-1552	79	14	and	and	CCONJ
ejpam-1552	79	15	me	i	PRON
ejpam-1552	79	16	models	model	NOUN
ejpam-1552	79	17	hold	hold	VERB
ejpam-1552	79	18	.	.	PUNCT
ejpam-1552	80	1	assume	assume	VERB
ejpam-1552	80	2	that	that	SCONJ
ejpam-1552	80	3	the	the	DET
ejpam-1552	80	4	c2rps	c2rps	PROPN
ejpam-1552	80	5	,	,	PUNCT
ejpam-1552	80	6	gs	gs	NOUN
ejpam-1552	80	7	,	,	PUNCT
ejpam-1552	80	8	and	and	CCONJ
ejpam-1552	80	9	me	i	PRON
ejpam-1552	80	10	models	model	NOUN
ejpam-1552	80	11	hold	hold	VERB
ejpam-1552	80	12	,	,	PUNCT
ejpam-1552	80	13	and	and	CCONJ
ejpam-1552	80	14	then	then	ADV
ejpam-1552	80	15	we	we	PRON
ejpam-1552	80	16	shall	shall	AUX
ejpam-1552	80	17	show	show	VERB
ejpam-1552	80	18	that	that	SCONJ
ejpam-1552	80	19	the	the	DET
ejpam-1552	80	20	s	s	PROPN
ejpam-1552	80	21	model	model	NOUN
ejpam-1552	80	22	holds	hold	VERB
ejpam-1552	80	23	.	.	PUNCT
ejpam-1552	81	1	we	we	PRON
ejpam-1552	81	2	see	see	VERB
ejpam-1552	81	3	e(x	e(x	NUM
ejpam-1552	81	4	)	)	PUNCT
ejpam-1552	82	1	=	=	PUNCT
ejpam-1552	82	2	r	r	NOUN
ejpam-1552	82	3	∑	∑	PROPN
ejpam-1552	82	4	i=1	i=1	PROPN
ejpam-1552	82	5	ipi	ipi	PROPN
ejpam-1552	82	6	·	·	PUNCT
ejpam-1552	82	7	k.	k.	PROPN
ejpam-1552	82	8	tahata	tahata	PROPN
ejpam-1552	82	9	,	,	PUNCT
ejpam-1552	82	10	k.	k.	PROPN
ejpam-1552	82	11	yamamoto	yamamoto	PROPN
ejpam-1552	82	12	,	,	PUNCT
ejpam-1552	82	13	s.	s.	PROPN
ejpam-1552	82	14	tomizawa	tomizawa	PROPN
ejpam-1552	82	15	/	/	SYM
ejpam-1552	82	16	eur	eur	PROPN
ejpam-1552	82	17	.	.	PUNCT
ejpam-1552	83	1	j.	j.	PROPN
ejpam-1552	83	2	pure	pure	PROPN
ejpam-1552	83	3	appl	appl	PROPN
ejpam-1552	83	4	.	.	PROPN
ejpam-1552	83	5	math	math	PROPN
ejpam-1552	83	6	,	,	PUNCT
ejpam-1552	83	7	6	6	NUM
ejpam-1552	83	8	(	(	PUNCT
ejpam-1552	83	9	2013	2013	NUM
ejpam-1552	83	10	)	)	PUNCT
ejpam-1552	83	11	,	,	PUNCT
ejpam-1552	83	12	299	299	NUM
ejpam-1552	83	13	-	-	SYM
ejpam-1552	83	14	306	306	NUM
ejpam-1552	83	15	302	302	NUM
ejpam-1552	83	16	=	=	NOUN
ejpam-1552	83	17	r	r	NOUN
ejpam-1552	83	18	∑	∑	PUNCT
ejpam-1552	83	19	s=1	s=1	ADP
ejpam-1552	83	20	r	r	NOUN
ejpam-1552	83	21	∑	∑	PROPN
ejpam-1552	83	22	t	t	PROPN
ejpam-1552	83	23	=	=	SYM
ejpam-1552	83	24	s	s	PART
ejpam-1552	83	25	pt	pt	X
ejpam-1552	83	26	·	·	PUNCT
ejpam-1552	83	27	=	=	PUNCT
ejpam-1552	83	28	r	r	NOUN
ejpam-1552	83	29	∑	∑	PUNCT
ejpam-1552	83	30	s=1	s=1	X
ejpam-1552	83	31	(	(	PUNCT
ejpam-1552	83	32	1−	1−	NUM
ejpam-1552	83	33	f	f	NOUN
ejpam-1552	83	34	x	x	SYM
ejpam-1552	83	35	s−1	s−1	PROPN
ejpam-1552	83	36	)	)	PUNCT
ejpam-1552	84	1	=	=	NOUN
ejpam-1552	84	2	r	r	NOUN
ejpam-1552	84	3	−	−	PROPN
ejpam-1552	84	4	r−1	r−1	PROPN
ejpam-1552	84	5	∑	∑	PROPN
ejpam-1552	84	6	i=1	i=1	PROPN
ejpam-1552	84	7	f	f	PROPN
ejpam-1552	84	8	x	x	VERB
ejpam-1552	85	1	i	i	PRON
ejpam-1552	85	2	,	,	PUNCT
ejpam-1552	85	3	where	where	SCONJ
ejpam-1552	85	4	f	f	NOUN
ejpam-1552	85	5	x	x	X
ejpam-1552	85	6	i	i	PROPN
ejpam-1552	85	7	=	=	PUNCT
ejpam-1552	85	8	p(x	p(x	VERB
ejpam-1552	85	9	≤	≤	NUM
ejpam-1552	85	10	i	i	PROPN
ejpam-1552	85	11	)	)	PUNCT
ejpam-1552	85	12	.	.	PUNCT
ejpam-1552	86	1	similarly	similarly	ADV
ejpam-1552	86	2	we	we	PRON
ejpam-1552	86	3	see	see	VERB
ejpam-1552	86	4	e(y	e(y	ADJ
ejpam-1552	86	5	)	)	PUNCT
ejpam-1552	87	1	=	=	PUNCT
ejpam-1552	88	1	r	r	NOUN
ejpam-1552	88	2	−	−	PROPN
ejpam-1552	88	3	r−1	r−1	PROPN
ejpam-1552	88	4	∑	∑	PUNCT
ejpam-1552	88	5	i=1	i=1	PROPN
ejpam-1552	88	6	f	f	PROPN
ejpam-1552	89	1	y	y	PROPN
ejpam-1552	90	1	i	i	PRON
ejpam-1552	90	2	,	,	PUNCT
ejpam-1552	90	3	where	where	SCONJ
ejpam-1552	90	4	f	f	PROPN
ejpam-1552	90	5	y	y	NOUN
ejpam-1552	90	6	i	i	PRON
ejpam-1552	90	7	=	=	PRON
ejpam-1552	90	8	p(y	p(y	VERB
ejpam-1552	90	9	≤	≤	NOUN
ejpam-1552	90	10	i	i	PROPN
ejpam-1552	90	11	)	)	PUNCT
ejpam-1552	90	12	.	.	PUNCT
ejpam-1552	91	1	thus	thus	ADV
ejpam-1552	91	2	we	we	PRON
ejpam-1552	91	3	see	see	VERB
ejpam-1552	91	4	e(y	e(y	ADJ
ejpam-1552	91	5	)	)	PUNCT
ejpam-1552	91	6	−	−	ADP
ejpam-1552	91	7	e(x	e(x	NUM
ejpam-1552	91	8	)	)	PUNCT
ejpam-1552	92	1	=	=	PUNCT
ejpam-1552	92	2	r−1	r−1	PROPN
ejpam-1552	92	3	∑	∑	PUNCT
ejpam-1552	92	4	i=1	i=1	PROPN
ejpam-1552	92	5	f	f	PROPN
ejpam-1552	93	1	x	x	INTJ
ejpam-1552	94	1	i	i	PRON
ejpam-1552	94	2	−	−	PROPN
ejpam-1552	95	1	r−1	r−1	PROPN
ejpam-1552	95	2	∑	∑	PROPN
ejpam-1552	95	3	i=1	i=1	PROPN
ejpam-1552	96	1	f	f	PROPN
ejpam-1552	96	2	y	y	PROPN
ejpam-1552	97	1	i	i	PRON
ejpam-1552	97	2	=	=	PUNCT
ejpam-1552	98	1	r−1	r−1	PROPN
ejpam-1552	98	2	∑	∑	PUNCT
ejpam-1552	98	3	i=1	i=1	PROPN
ejpam-1552	98	4	gi	gi	INTJ
ejpam-1552	98	5	,	,	PUNCT
ejpam-1552	98	6	i+1−	i+1−	PROPN
ejpam-1552	98	7	r−1	r−1	PROPN
ejpam-1552	98	8	∑	∑	PUNCT
ejpam-1552	98	9	i=1	i=1	PROPN
ejpam-1552	98	10	gi+1,i	gi+1,i	PROPN
ejpam-1552	98	11	.	.	PUNCT
ejpam-1552	99	1	from	from	ADP
ejpam-1552	99	2	the	the	DET
ejpam-1552	99	3	me	me	PROPN
ejpam-1552	99	4	model	model	NOUN
ejpam-1552	99	5	,	,	PUNCT
ejpam-1552	99	6	we	we	PRON
ejpam-1552	99	7	see	see	VERB
ejpam-1552	99	8	r−1	r−1	PROPN
ejpam-1552	99	9	∑	∑	PROPN
ejpam-1552	99	10	i=1	i=1	PROPN
ejpam-1552	99	11	gi	gi	NOUN
ejpam-1552	99	12	,	,	PUNCT
ejpam-1552	99	13	i+1	i+1	NOUN
ejpam-1552	99	14	=	=	SYM
ejpam-1552	99	15	r−1	r−1	PROPN
ejpam-1552	99	16	∑	∑	PUNCT
ejpam-1552	99	17	i=1	i=1	PROPN
ejpam-1552	99	18	gi+1,i	gi+1,i	PROPN
ejpam-1552	99	19	.	.	PUNCT
ejpam-1552	100	1	(	(	PUNCT
ejpam-1552	100	2	1	1	X
ejpam-1552	100	3	)	)	PUNCT
ejpam-1552	100	4	from	from	ADP
ejpam-1552	100	5	the	the	DET
ejpam-1552	100	6	c2rps	c2rps	PROPN
ejpam-1552	100	7	model	model	NOUN
ejpam-1552	100	8	,	,	PUNCT
ejpam-1552	100	9	we	we	PRON
ejpam-1552	100	10	obtain	obtain	VERB
ejpam-1552	100	11	r−1	r−1	PROPN
ejpam-1552	100	12	∑	∑	PROPN
ejpam-1552	100	13	i=1	i=1	PROPN
ejpam-1552	100	14	gi	gi	NOUN
ejpam-1552	100	15	,	,	PUNCT
ejpam-1552	100	16	i+1	i+1	ADV
ejpam-1552	100	17	=	=	PUNCT
ejpam-1552	100	18	γθ	γθ	PROPN
ejpam-1552	100	19	r−1	r−1	PROPN
ejpam-1552	100	20	∑	∑	PUNCT
ejpam-1552	100	21	i=1	i=1	PROPN
ejpam-1552	100	22	gi+1,i	gi+1,i	PROPN
ejpam-1552	100	23	.	.	PUNCT
ejpam-1552	101	1	from	from	ADP
ejpam-1552	101	2	(	(	PUNCT
ejpam-1552	101	3	1	1	X
ejpam-1552	101	4	)	)	PUNCT
ejpam-1552	101	5	we	we	PRON
ejpam-1552	101	6	see	see	VERB
ejpam-1552	101	7	γ	γ	X
ejpam-1552	101	8	=	=	SYM
ejpam-1552	101	9	θ−1	θ−1	PROPN
ejpam-1552	101	10	.	.	PUNCT
ejpam-1552	102	1	thus	thus	ADV
ejpam-1552	102	2	gi	gi	VERB
ejpam-1552	102	3	j	j	PROPN
ejpam-1552	102	4	g	g	PROPN
ejpam-1552	102	5	ji	ji	PROPN
ejpam-1552	103	1	=	=	PROPN
ejpam-1552	103	2	θ	θ	X
ejpam-1552	103	3	j−i−1	j−i−1	PROPN
ejpam-1552	103	4	(	(	PUNCT
ejpam-1552	103	5	i	i	PRON
ejpam-1552	103	6	<	<	X
ejpam-1552	103	7	j	j	PROPN
ejpam-1552	103	8	)	)	PUNCT
ejpam-1552	103	9	.	.	PUNCT
ejpam-1552	104	1	(	(	PUNCT
ejpam-1552	104	2	2	2	X
ejpam-1552	104	3	)	)	PUNCT
ejpam-1552	104	4	we	we	PRON
ejpam-1552	104	5	can	can	AUX
ejpam-1552	104	6	see	see	VERB
ejpam-1552	104	7	that	that	SCONJ
ejpam-1552	104	8	r−1	r−1	PROPN
ejpam-1552	104	9	∑	∑	PROPN
ejpam-1552	104	10	i=1	i=1	PROPN
ejpam-1552	104	11	gi	gi	NOUN
ejpam-1552	104	12	,	,	PUNCT
ejpam-1552	104	13	i+1	i+1	NOUN
ejpam-1552	105	1	=	=	SYM
ejpam-1552	105	2	r−1	r−1	PROPN
ejpam-1552	105	3	∑	∑	PUNCT
ejpam-1552	105	4	i=1	i=1	PROPN
ejpam-1552	105	5	r	r	PROPN
ejpam-1552	105	6	∑	∑	PUNCT
ejpam-1552	105	7	j	j	X
ejpam-1552	105	8	=	=	PROPN
ejpam-1552	105	9	i+1	i+1	X
ejpam-1552	105	10	(	(	PUNCT
ejpam-1552	105	11	j−	j−	VERB
ejpam-1552	105	12	i)pi	i)pi	PROPN
ejpam-1552	105	13	j	j	PROPN
ejpam-1552	105	14	,	,	PUNCT
ejpam-1552	105	15	and	and	CCONJ
ejpam-1552	105	16	r−1	r−1	PROPN
ejpam-1552	105	17	∑	∑	PUNCT
ejpam-1552	106	1	i=1	i=1	PROPN
ejpam-1552	106	2	gi+1,i	gi+1,i	NOUN
ejpam-1552	106	3	=	=	PUNCT
ejpam-1552	107	1	r−1	r−1	PROPN
ejpam-1552	107	2	∑	∑	PUNCT
ejpam-1552	107	3	i=1	i=1	PROPN
ejpam-1552	107	4	r	r	PROPN
ejpam-1552	107	5	∑	∑	PUNCT
ejpam-1552	107	6	j	j	X
ejpam-1552	107	7	=	=	PROPN
ejpam-1552	107	8	i+1	i+1	X
ejpam-1552	107	9	(	(	PUNCT
ejpam-1552	107	10	j−	j−	VERB
ejpam-1552	107	11	i)p	i)p	ADJ
ejpam-1552	107	12	ji	ji	PROPN
ejpam-1552	107	13	.	.	PUNCT
ejpam-1552	108	1	k.	k.	PROPN
ejpam-1552	108	2	tahata	tahata	PROPN
ejpam-1552	108	3	,	,	PUNCT
ejpam-1552	108	4	k.	k.	PROPN
ejpam-1552	108	5	yamamoto	yamamoto	PROPN
ejpam-1552	108	6	,	,	PUNCT
ejpam-1552	108	7	s.	s.	PROPN
ejpam-1552	108	8	tomizawa	tomizawa	PROPN
ejpam-1552	108	9	/	/	SYM
ejpam-1552	108	10	eur	eur	PROPN
ejpam-1552	108	11	.	.	PUNCT
ejpam-1552	109	1	j.	j.	PROPN
ejpam-1552	109	2	pure	pure	PROPN
ejpam-1552	109	3	appl	appl	PROPN
ejpam-1552	109	4	.	.	PROPN
ejpam-1552	109	5	math	math	PROPN
ejpam-1552	109	6	,	,	PUNCT
ejpam-1552	109	7	6	6	NUM
ejpam-1552	109	8	(	(	PUNCT
ejpam-1552	109	9	2013	2013	NUM
ejpam-1552	109	10	)	)	PUNCT
ejpam-1552	109	11	,	,	PUNCT
ejpam-1552	109	12	299	299	NUM
ejpam-1552	109	13	-	-	SYM
ejpam-1552	109	14	306	306	NUM
ejpam-1552	109	15	303	303	NUM
ejpam-1552	109	16	also	also	ADV
ejpam-1552	109	17	we	we	PRON
ejpam-1552	109	18	can	can	AUX
ejpam-1552	109	19	see	see	VERB
ejpam-1552	109	20	that	that	SCONJ
ejpam-1552	109	21	r−2	r−2	PROPN
ejpam-1552	109	22	∑	∑	PUNCT
ejpam-1552	109	23	i=1	i=1	PROPN
ejpam-1552	109	24	gi	gi	NOUN
ejpam-1552	109	25	,	,	PUNCT
ejpam-1552	109	26	i+2	i+2	PROPN
ejpam-1552	109	27	=	=	PUNCT
ejpam-1552	109	28	r−2	r−2	PROPN
ejpam-1552	109	29	∑	∑	PUNCT
ejpam-1552	109	30	i=1	i=1	PROPN
ejpam-1552	109	31	r	r	NOUN
ejpam-1552	109	32	∑	∑	PUNCT
ejpam-1552	109	33	j	j	X
ejpam-1552	109	34	=	=	AUX
ejpam-1552	109	35	i+2	i+2	PROPN
ejpam-1552	109	36	(	(	PUNCT
ejpam-1552	109	37	j−	j−	VERB
ejpam-1552	109	38	i−	i−	PROPN
ejpam-1552	109	39	1)pi	1)pi	PROPN
ejpam-1552	109	40	j	j	PROPN
ejpam-1552	109	41	,	,	PUNCT
ejpam-1552	109	42	and	and	CCONJ
ejpam-1552	109	43	r−2	r−2	PROPN
ejpam-1552	109	44	∑	∑	PROPN
ejpam-1552	109	45	i=1	i=1	PROPN
ejpam-1552	109	46	gi+2,i	gi+2,i	PROPN
ejpam-1552	109	47	=	=	SYM
ejpam-1552	110	1	r−2	r−2	PROPN
ejpam-1552	110	2	∑	∑	PUNCT
ejpam-1552	110	3	i=1	i=1	PROPN
ejpam-1552	110	4	r	r	NOUN
ejpam-1552	110	5	∑	∑	PUNCT
ejpam-1552	110	6	j	j	X
ejpam-1552	110	7	=	=	AUX
ejpam-1552	110	8	i+2	i+2	PROPN
ejpam-1552	110	9	(	(	PUNCT
ejpam-1552	110	10	j−	j−	PROPN
ejpam-1552	110	11	i−	i−	PROPN
ejpam-1552	111	1	1)p	1)p	NUM
ejpam-1552	111	2	ji	ji	INTJ
ejpam-1552	111	3	.	.	PUNCT
ejpam-1552	112	1	therefore	therefore	ADV
ejpam-1552	112	2	we	we	PRON
ejpam-1552	112	3	can	can	AUX
ejpam-1552	112	4	see	see	VERB
ejpam-1552	112	5	that	that	SCONJ
ejpam-1552	112	6	δu	δu	ADP
ejpam-1552	112	7	=	=	PUNCT
ejpam-1552	112	8	r−1	r−1	PROPN
ejpam-1552	112	9	∑	∑	PROPN
ejpam-1552	112	10	i=1	i=1	PROPN
ejpam-1552	112	11	gi	gi	INTJ
ejpam-1552	112	12	,	,	PUNCT
ejpam-1552	112	13	i+1−	i+1−	PROPN
ejpam-1552	112	14	r−2	r−2	PROPN
ejpam-1552	112	15	∑	∑	PROPN
ejpam-1552	112	16	i=1	i=1	PROPN
ejpam-1552	112	17	gi	gi	PROPN
ejpam-1552	112	18	,	,	PUNCT
ejpam-1552	112	19	i+2	i+2	PROPN
ejpam-1552	112	20	,	,	PUNCT
ejpam-1552	112	21	and	and	CCONJ
ejpam-1552	112	22	δl	δl	X
ejpam-1552	112	23	=	=	SYM
ejpam-1552	112	24	r−1	r−1	PROPN
ejpam-1552	112	25	∑	∑	PUNCT
ejpam-1552	112	26	i=1	i=1	PROPN
ejpam-1552	112	27	gi+1,i	gi+1,i	PROPN
ejpam-1552	112	28	−	−	PROPN
ejpam-1552	112	29	r−2	r−2	PROPN
ejpam-1552	112	30	∑	∑	PROPN
ejpam-1552	112	31	i=1	i=1	PROPN
ejpam-1552	112	32	gi+2,i	gi+2,i	PROPN
ejpam-1552	112	33	.	.	PUNCT
ejpam-1552	113	1	from	from	ADP
ejpam-1552	113	2	the	the	DET
ejpam-1552	113	3	gs	gs	PROPN
ejpam-1552	113	4	model	model	NOUN
ejpam-1552	113	5	(	(	PUNCT
ejpam-1552	113	6	i.e.	i.e.	X
ejpam-1552	113	7	,	,	PUNCT
ejpam-1552	113	8	δu	δu	X
ejpam-1552	113	9	=	=	SYM
ejpam-1552	113	10	δl	δl	X
ejpam-1552	113	11	)	)	PUNCT
ejpam-1552	113	12	and	and	CCONJ
ejpam-1552	113	13	from	from	ADP
ejpam-1552	113	14	(	(	PUNCT
ejpam-1552	113	15	1	1	NUM
ejpam-1552	113	16	)	)	PUNCT
ejpam-1552	113	17	,	,	PUNCT
ejpam-1552	113	18	we	we	PRON
ejpam-1552	113	19	can	can	AUX
ejpam-1552	113	20	obtain	obtain	VERB
ejpam-1552	113	21	r−2	r−2	PROPN
ejpam-1552	113	22	∑	∑	PUNCT
ejpam-1552	113	23	i=1	i=1	PROPN
ejpam-1552	113	24	gi	gi	NOUN
ejpam-1552	113	25	,	,	PUNCT
ejpam-1552	113	26	i+2	i+2	PROPN
ejpam-1552	114	1	=	=	PUNCT
ejpam-1552	114	2	r−2	r−2	PROPN
ejpam-1552	114	3	∑	∑	PUNCT
ejpam-1552	114	4	i=1	i=1	PROPN
ejpam-1552	114	5	gi+2,i	gi+2,i	PROPN
ejpam-1552	114	6	.	.	PUNCT
ejpam-1552	115	1	from	from	ADP
ejpam-1552	115	2	(	(	PUNCT
ejpam-1552	115	3	2	2	X
ejpam-1552	115	4	)	)	PUNCT
ejpam-1552	115	5	we	we	PRON
ejpam-1552	115	6	obtain	obtain	VERB
ejpam-1552	115	7	r−2	r−2	PROPN
ejpam-1552	115	8	∑	∑	PUNCT
ejpam-1552	115	9	i=1	i=1	PROPN
ejpam-1552	115	10	θgi+2,i	θgi+2,i	PROPN
ejpam-1552	115	11	=	=	SYM
ejpam-1552	115	12	r−2	r−2	PROPN
ejpam-1552	115	13	∑	∑	PUNCT
ejpam-1552	115	14	i=1	i=1	PROPN
ejpam-1552	115	15	gi+2,i	gi+2,i	PROPN
ejpam-1552	115	16	.	.	PUNCT
ejpam-1552	116	1	thus	thus	ADV
ejpam-1552	116	2	θ=	θ=	PROPN
ejpam-1552	116	3	1	1	NUM
ejpam-1552	116	4	,	,	PUNCT
ejpam-1552	116	5	i.e.	i.e.	X
ejpam-1552	116	6	,	,	PUNCT
ejpam-1552	116	7	the	the	DET
ejpam-1552	116	8	s	s	NOUN
ejpam-1552	116	9	model	model	NOUN
ejpam-1552	116	10	holds	hold	VERB
ejpam-1552	116	11	.	.	PUNCT
ejpam-1552	117	1	the	the	DET
ejpam-1552	117	2	proof	proof	NOUN
ejpam-1552	117	3	is	be	AUX
ejpam-1552	117	4	completed	complete	VERB
ejpam-1552	117	5	.	.	PUNCT
ejpam-1552	118	1	3	3	X
ejpam-1552	118	2	.	.	X
ejpam-1552	118	3	goodness	goodness	NOUN
ejpam-1552	118	4	-	-	PUNCT
ejpam-1552	118	5	of	of	ADP
ejpam-1552	118	6	-	-	PUNCT
ejpam-1552	118	7	fit	fit	ADJ
ejpam-1552	118	8	test	test	NOUN
ejpam-1552	118	9	let	let	VERB
ejpam-1552	118	10	x	x	PUNCT
ejpam-1552	118	11	i	i	PRON
ejpam-1552	118	12	j	j	PROPN
ejpam-1552	118	13	denote	denote	VERB
ejpam-1552	118	14	the	the	DET
ejpam-1552	118	15	observed	observe	VERB
ejpam-1552	118	16	frequency	frequency	NOUN
ejpam-1552	118	17	in	in	ADP
ejpam-1552	118	18	the	the	DET
ejpam-1552	118	19	ith	ith	PROPN
ejpam-1552	118	20	row	row	NOUN
ejpam-1552	118	21	and	and	CCONJ
ejpam-1552	118	22	jth	jth	PROPN
ejpam-1552	118	23	column	column	NOUN
ejpam-1552	118	24	of	of	ADP
ejpam-1552	118	25	the	the	DET
ejpam-1552	118	26	r	r	NOUN
ejpam-1552	118	27	×	×	NOUN
ejpam-1552	118	28	r	r	NOUN
ejpam-1552	118	29	table	table	NOUN
ejpam-1552	118	30	(	(	PUNCT
ejpam-1552	118	31	i	i	NOUN
ejpam-1552	118	32	=	=	NOUN
ejpam-1552	118	33	1	1	NUM
ejpam-1552	118	34	,	,	PUNCT
ejpam-1552	118	35	.	.	PUNCT
ejpam-1552	118	36	.	.	PUNCT
ejpam-1552	119	1	.	.	PUNCT
ejpam-1552	120	1	,	,	PUNCT
ejpam-1552	120	2	r	r	NOUN
ejpam-1552	120	3	;	;	PUNCT
ejpam-1552	120	4	j	j	PROPN
ejpam-1552	120	5	=	=	SYM
ejpam-1552	120	6	1	1	NUM
ejpam-1552	120	7	,	,	PUNCT
ejpam-1552	120	8	.	.	PUNCT
ejpam-1552	120	9	.	.	PUNCT
ejpam-1552	120	10	.	.	PUNCT
ejpam-1552	121	1	,	,	PUNCT
ejpam-1552	121	2	r	r	NOUN
ejpam-1552	121	3	)	)	PUNCT
ejpam-1552	121	4	,	,	PUNCT
ejpam-1552	121	5	with	with	ADP
ejpam-1552	121	6	n	n	NOUN
ejpam-1552	121	7	=	=	SYM
ejpam-1552	122	1	∑∑	∑∑	PROPN
ejpam-1552	122	2	x	x	PUNCT
ejpam-1552	123	1	i	i	PRON
ejpam-1552	123	2	j	j	PROPN
ejpam-1552	123	3	.	.	PUNCT
ejpam-1552	124	1	let	let	VERB
ejpam-1552	124	2	mi	mi	PROPN
ejpam-1552	124	3	j	j	PROPN
ejpam-1552	124	4	denote	denote	VERB
ejpam-1552	124	5	the	the	DET
ejpam-1552	124	6	corresponding	corresponding	ADJ
ejpam-1552	124	7	expected	expect	VERB
ejpam-1552	124	8	frequency	frequency	NOUN
ejpam-1552	124	9	.	.	PUNCT
ejpam-1552	125	1	assuming	assume	VERB
ejpam-1552	125	2	that	that	SCONJ
ejpam-1552	125	3	{	{	PUNCT
ejpam-1552	125	4	x	x	SYM
ejpam-1552	125	5	i	i	PRON
ejpam-1552	125	6	j	j	PROPN
ejpam-1552	125	7	}	}	PUNCT
ejpam-1552	125	8	have	have	VERB
ejpam-1552	125	9	a	a	DET
ejpam-1552	125	10	multinomial	multinomial	ADJ
ejpam-1552	125	11	distribution	distribution	NOUN
ejpam-1552	125	12	,	,	PUNCT
ejpam-1552	125	13	the	the	DET
ejpam-1552	125	14	maximum	maximum	ADJ
ejpam-1552	125	15	likelihood	likelihood	NOUN
ejpam-1552	125	16	estimates	estimate	NOUN
ejpam-1552	125	17	of	of	ADP
ejpam-1552	125	18	expected	expect	VERB
ejpam-1552	125	19	frequencies	frequency	NOUN
ejpam-1552	125	20	{	{	PUNCT
ejpam-1552	125	21	mi	mi	PROPN
ejpam-1552	125	22	j	j	PROPN
ejpam-1552	125	23	}	}	PUNCT
ejpam-1552	125	24	under	under	ADP
ejpam-1552	125	25	each	each	DET
ejpam-1552	125	26	model	model	NOUN
ejpam-1552	125	27	could	could	AUX
ejpam-1552	125	28	be	be	AUX
ejpam-1552	125	29	obtained	obtain	VERB
ejpam-1552	125	30	,	,	PUNCT
ejpam-1552	125	31	for	for	ADP
ejpam-1552	125	32	example	example	NOUN
ejpam-1552	125	33	,	,	PUNCT
ejpam-1552	125	34	using	use	VERB
ejpam-1552	125	35	the	the	DET
ejpam-1552	125	36	newton	newton	PROPN
ejpam-1552	125	37	-	-	PUNCT
ejpam-1552	125	38	raphson	raphson	PROPN
ejpam-1552	125	39	method	method	NOUN
ejpam-1552	125	40	to	to	ADP
ejpam-1552	125	41	the	the	DET
ejpam-1552	125	42	log	log	NOUN
ejpam-1552	125	43	-	-	PUNCT
ejpam-1552	125	44	likelihood	likelihood	NOUN
ejpam-1552	125	45	equations	equation	NOUN
ejpam-1552	125	46	.	.	PUNCT
ejpam-1552	126	1	the	the	DET
ejpam-1552	126	2	goodness	goodness	NOUN
ejpam-1552	126	3	-	-	PUNCT
ejpam-1552	126	4	of	of	ADP
ejpam-1552	126	5	-	-	PUNCT
ejpam-1552	126	6	fit	fit	NOUN
ejpam-1552	126	7	of	of	ADP
ejpam-1552	126	8	each	each	DET
ejpam-1552	126	9	model	model	NOUN
ejpam-1552	126	10	can	can	AUX
ejpam-1552	126	11	be	be	AUX
ejpam-1552	126	12	tested	test	VERB
ejpam-1552	126	13	by	by	ADP
ejpam-1552	126	14	,	,	PUNCT
ejpam-1552	126	15	e.g.	e.g.	ADV
ejpam-1552	126	16	,	,	PUNCT
ejpam-1552	126	17	the	the	DET
ejpam-1552	126	18	likelihood	likelihood	NOUN
ejpam-1552	126	19	ratio	ratio	NOUN
ejpam-1552	126	20	chi	chi	NOUN
ejpam-1552	126	21	-	-	PUNCT
ejpam-1552	126	22	squared	square	VERB
ejpam-1552	126	23	statistic	statistic	NOUN
ejpam-1552	126	24	g2	g2	PROPN
ejpam-1552	126	25	with	with	ADP
ejpam-1552	126	26	the	the	DET
ejpam-1552	126	27	corresponding	corresponding	ADJ
ejpam-1552	126	28	degrees	degree	NOUN
ejpam-1552	126	29	of	of	ADP
ejpam-1552	126	30	freedom	freedom	NOUN
ejpam-1552	126	31	,	,	PUNCT
ejpam-1552	126	32	defined	define	VERB
ejpam-1552	126	33	by	by	ADP
ejpam-1552	126	34	g2	g2	PROPN
ejpam-1552	126	35	=	=	SYM
ejpam-1552	126	36	2	2	NUM
ejpam-1552	126	37	r	r	NOUN
ejpam-1552	126	38	∑	∑	NOUN
ejpam-1552	126	39	i=1	i=1	INTJ
ejpam-1552	126	40	r	r	NOUN
ejpam-1552	126	41	∑	∑	PUNCT
ejpam-1552	126	42	j=1	j=1	NOUN
ejpam-1552	126	43	x	x	PUNCT
ejpam-1552	127	1	i	i	PRON
ejpam-1552	127	2	j	j	PROPN
ejpam-1552	127	3	log	log	VERB
ejpam-1552	127	4	�	�	PROPN
ejpam-1552	127	5	x	x	PUNCT
ejpam-1552	128	1	i	i	PRON
ejpam-1552	128	2	j	j	PROPN
ejpam-1552	128	3	m̂i	m̂i	PROPN
ejpam-1552	128	4	j	j	PROPN
ejpam-1552	128	5	�	�	PROPN
ejpam-1552	128	6	,	,	PUNCT
ejpam-1552	128	7	where	where	SCONJ
ejpam-1552	128	8	m̂i	m̂i	NOUN
ejpam-1552	128	9	j	j	PROPN
ejpam-1552	128	10	is	be	AUX
ejpam-1552	128	11	the	the	DET
ejpam-1552	128	12	maximum	maximum	ADJ
ejpam-1552	128	13	likelihood	likelihood	NOUN
ejpam-1552	128	14	estimate	estimate	NOUN
ejpam-1552	128	15	of	of	ADP
ejpam-1552	128	16	mi	mi	PROPN
ejpam-1552	128	17	j	j	PROPN
ejpam-1552	128	18	under	under	ADP
ejpam-1552	128	19	the	the	DET
ejpam-1552	128	20	model	model	NOUN
ejpam-1552	128	21	.	.	PUNCT
ejpam-1552	129	1	the	the	DET
ejpam-1552	129	2	numbers	number	NOUN
ejpam-1552	129	3	of	of	ADP
ejpam-1552	129	4	degrees	degree	NOUN
ejpam-1552	129	5	of	of	ADP
ejpam-1552	129	6	freedom	freedom	NOUN
ejpam-1552	129	7	for	for	ADP
ejpam-1552	129	8	each	each	DET
ejpam-1552	129	9	model	model	NOUN
ejpam-1552	129	10	are	be	AUX
ejpam-1552	129	11	omitted	omit	VERB
ejpam-1552	129	12	here	here	ADV
ejpam-1552	129	13	;	;	PUNCT
ejpam-1552	129	14	however	however	ADV
ejpam-1552	129	15	,	,	PUNCT
ejpam-1552	129	16	when	when	SCONJ
ejpam-1552	129	17	r	r	NOUN
ejpam-1552	129	18	=	=	SYM
ejpam-1552	129	19	4	4	NUM
ejpam-1552	129	20	,	,	PUNCT
ejpam-1552	129	21	those	those	PRON
ejpam-1552	129	22	are	be	AUX
ejpam-1552	129	23	given	give	VERB
ejpam-1552	129	24	in	in	ADP
ejpam-1552	129	25	table	table	NOUN
ejpam-1552	129	26	2	2	NUM
ejpam-1552	129	27	.	.	PUNCT
ejpam-1552	129	28	k.	k.	PROPN
ejpam-1552	129	29	tahata	tahata	PROPN
ejpam-1552	129	30	,	,	PUNCT
ejpam-1552	129	31	k.	k.	PROPN
ejpam-1552	129	32	yamamoto	yamamoto	PROPN
ejpam-1552	129	33	,	,	PUNCT
ejpam-1552	129	34	s.	s.	PROPN
ejpam-1552	129	35	tomizawa	tomizawa	PROPN
ejpam-1552	129	36	/	/	SYM
ejpam-1552	129	37	eur	eur	PROPN
ejpam-1552	129	38	.	.	PUNCT
ejpam-1552	130	1	j.	j.	PROPN
ejpam-1552	130	2	pure	pure	PROPN
ejpam-1552	130	3	appl	appl	PROPN
ejpam-1552	130	4	.	.	PROPN
ejpam-1552	130	5	math	math	PROPN
ejpam-1552	130	6	,	,	PUNCT
ejpam-1552	130	7	6	6	NUM
ejpam-1552	130	8	(	(	PUNCT
ejpam-1552	130	9	2013	2013	NUM
ejpam-1552	130	10	)	)	PUNCT
ejpam-1552	130	11	,	,	PUNCT
ejpam-1552	130	12	299	299	NUM
ejpam-1552	130	13	-	-	SYM
ejpam-1552	130	14	306	306	NUM
ejpam-1552	130	15	304	304	NUM
ejpam-1552	130	16	4	4	NUM
ejpam-1552	130	17	.	.	PUNCT
ejpam-1552	130	18	example	example	NOUN
ejpam-1552	130	19	consider	consider	VERB
ejpam-1552	130	20	the	the	DET
ejpam-1552	130	21	vision	vision	NOUN
ejpam-1552	130	22	data	datum	NOUN
ejpam-1552	130	23	in	in	ADP
ejpam-1552	130	24	table	table	NOUN
ejpam-1552	130	25	1	1	NUM
ejpam-1552	130	26	.	.	PUNCT
ejpam-1552	131	1	the	the	DET
ejpam-1552	131	2	row	row	NOUN
ejpam-1552	131	3	variable	variable	NOUN
ejpam-1552	131	4	x	x	PUNCT
ejpam-1552	131	5	is	be	AUX
ejpam-1552	131	6	the	the	DET
ejpam-1552	131	7	right	right	ADJ
ejpam-1552	131	8	eye	eye	NOUN
ejpam-1552	131	9	grade	grade	NOUN
ejpam-1552	131	10	and	and	CCONJ
ejpam-1552	131	11	the	the	DET
ejpam-1552	131	12	column	column	NOUN
ejpam-1552	131	13	variable	variable	NOUN
ejpam-1552	131	14	y	y	PROPN
ejpam-1552	131	15	is	be	AUX
ejpam-1552	131	16	the	the	DET
ejpam-1552	131	17	left	left	ADJ
ejpam-1552	131	18	eye	eye	NOUN
ejpam-1552	131	19	grade	grade	NOUN
ejpam-1552	131	20	.	.	PUNCT
ejpam-1552	132	1	the	the	DET
ejpam-1552	132	2	categories	category	NOUN
ejpam-1552	132	3	are	be	AUX
ejpam-1552	132	4	ordered	order	VERB
ejpam-1552	132	5	from	from	ADP
ejpam-1552	132	6	best	well	ADV
ejpam-1552	132	7	(	(	PUNCT
ejpam-1552	132	8	1	1	NUM
ejpam-1552	132	9	)	)	PUNCT
ejpam-1552	132	10	to	to	PART
ejpam-1552	132	11	worst	worst	ADV
ejpam-1552	132	12	(	(	PUNCT
ejpam-1552	132	13	4	4	NUM
ejpam-1552	132	14	)	)	PUNCT
ejpam-1552	132	15	.	.	PUNCT
ejpam-1552	133	1	these	these	DET
ejpam-1552	133	2	data	datum	NOUN
ejpam-1552	133	3	have	have	AUX
ejpam-1552	133	4	been	be	AUX
ejpam-1552	133	5	analyzed	analyze	VERB
ejpam-1552	133	6	by	by	ADP
ejpam-1552	133	7	many	many	ADJ
ejpam-1552	133	8	statisticians	statistician	NOUN
ejpam-1552	133	9	,	,	PUNCT
ejpam-1552	133	10	including	include	VERB
ejpam-1552	133	11	stuart	stuart	PROPN
ejpam-1552	133	12	[	[	X
ejpam-1552	133	13	8	8	NUM
ejpam-1552	133	14	]	]	PUNCT
ejpam-1552	133	15	,	,	PUNCT
ejpam-1552	133	16	bishop	bishop	PROPN
ejpam-1552	133	17	et	et	PROPN
ejpam-1552	133	18	al	al	PROPN
ejpam-1552	133	19	.	.	PUNCT
ejpam-1552	134	1	[	[	X
ejpam-1552	134	2	2	2	NUM
ejpam-1552	134	3	,	,	PUNCT
ejpam-1552	134	4	p.284	p.284	NOUN
ejpam-1552	134	5	]	]	X
ejpam-1552	134	6	,	,	PUNCT
ejpam-1552	134	7	mccullagh	mccullagh	NOUN
ejpam-1552	135	1	[	[	X
ejpam-1552	135	2	6	6	NUM
ejpam-1552	135	3	]	]	PUNCT
ejpam-1552	135	4	,	,	PUNCT
ejpam-1552	135	5	goodman	goodman	PROPN
ejpam-1552	136	1	[	[	X
ejpam-1552	136	2	5	5	NUM
ejpam-1552	136	3	]	]	PUNCT
ejpam-1552	136	4	,	,	PUNCT
ejpam-1552	136	5	tomizawa	tomizawa	PROPN
ejpam-1552	137	1	[	[	X
ejpam-1552	137	2	12	12	NUM
ejpam-1552	137	3	]	]	PUNCT
ejpam-1552	137	4	,	,	PUNCT
ejpam-1552	137	5	miyamoto	miyamoto	PROPN
ejpam-1552	137	6	et	et	PROPN
ejpam-1552	137	7	al	al	PROPN
ejpam-1552	137	8	.	.	PUNCT
ejpam-1552	138	1	[	[	X
ejpam-1552	138	2	7	7	NUM
ejpam-1552	138	3	]	]	PUNCT
ejpam-1552	138	4	,	,	PUNCT
ejpam-1552	138	5	and	and	CCONJ
ejpam-1552	138	6	tomizawa	tomizawa	PROPN
ejpam-1552	138	7	and	and	CCONJ
ejpam-1552	138	8	tahata	tahata	VERB
ejpam-1552	139	1	[	[	X
ejpam-1552	139	2	14	14	NUM
ejpam-1552	139	3	]	]	PUNCT
ejpam-1552	139	4	.	.	PUNCT
ejpam-1552	140	1	table	table	NOUN
ejpam-1552	140	2	1	1	NUM
ejpam-1552	140	3	:	:	PUNCT
ejpam-1552	140	4	unaided	unaided	ADJ
ejpam-1552	140	5	distance	distance	NOUN
ejpam-1552	140	6	vision	vision	NOUN
ejpam-1552	140	7	of	of	ADP
ejpam-1552	140	8	7	7	NUM
ejpam-1552	140	9	,	,	PUNCT
ejpam-1552	140	10	477	477	NUM
ejpam-1552	140	11	women	woman	NOUN
ejpam-1552	140	12	aged	age	VERB
ejpam-1552	140	13	30	30	NUM
ejpam-1552	140	14	-	-	SYM
ejpam-1552	140	15	39	39	NUM
ejpam-1552	140	16	employed	employ	VERB
ejpam-1552	140	17	in	in	ADP
ejpam-1552	140	18	royal	royal	ADJ
ejpam-1552	140	19	ordnance	ordnance	NOUN
ejpam-1552	140	20	factories	factory	NOUN
ejpam-1552	140	21	in	in	ADP
ejpam-1552	140	22	britain	britain	PROPN
ejpam-1552	140	23	from	from	ADP
ejpam-1552	140	24	1943	1943	NUM
ejpam-1552	140	25	to	to	ADP
ejpam-1552	140	26	1946	1946	NUM
ejpam-1552	140	27	;	;	PUNCT
ejpam-1552	140	28	from	from	ADP
ejpam-1552	140	29	[	[	X
ejpam-1552	140	30	8	8	NUM
ejpam-1552	140	31	]	]	PUNCT
ejpam-1552	140	32	.	.	PUNCT
ejpam-1552	141	1	right	right	ADJ
ejpam-1552	141	2	eye	eye	NOUN
ejpam-1552	141	3	left	leave	VERB
ejpam-1552	141	4	eye	eye	NOUN
ejpam-1552	141	5	grade	grade	NOUN
ejpam-1552	141	6	grade	grade	NOUN
ejpam-1552	141	7	best	well	ADV
ejpam-1552	141	8	(	(	PUNCT
ejpam-1552	141	9	1	1	NUM
ejpam-1552	141	10	)	)	PUNCT
ejpam-1552	141	11	second	second	ADJ
ejpam-1552	141	12	(	(	PUNCT
ejpam-1552	141	13	2	2	NUM
ejpam-1552	141	14	)	)	PUNCT
ejpam-1552	141	15	third	third	NOUN
ejpam-1552	141	16	(	(	PUNCT
ejpam-1552	141	17	3	3	X
ejpam-1552	141	18	)	)	PUNCT
ejpam-1552	141	19	worst	bad	ADJ
ejpam-1552	141	20	(	(	PUNCT
ejpam-1552	141	21	4	4	X
ejpam-1552	141	22	)	)	PUNCT
ejpam-1552	141	23	total	total	NOUN
ejpam-1552	141	24	best	good	ADJ
ejpam-1552	141	25	(	(	PUNCT
ejpam-1552	141	26	1	1	X
ejpam-1552	141	27	)	)	PUNCT
ejpam-1552	141	28	1520	1520	NUM
ejpam-1552	141	29	266	266	NUM
ejpam-1552	141	30	124	124	NUM
ejpam-1552	141	31	66	66	NUM
ejpam-1552	141	32	1976	1976	NUM
ejpam-1552	141	33	second	second	ADJ
ejpam-1552	141	34	(	(	PUNCT
ejpam-1552	141	35	2	2	NUM
ejpam-1552	141	36	)	)	PUNCT
ejpam-1552	141	37	234	234	NUM
ejpam-1552	141	38	1512	1512	NUM
ejpam-1552	141	39	432	432	NUM
ejpam-1552	141	40	78	78	NUM
ejpam-1552	141	41	2256	2256	NUM
ejpam-1552	141	42	third	third	ADJ
ejpam-1552	141	43	(	(	PUNCT
ejpam-1552	141	44	3	3	NUM
ejpam-1552	141	45	)	)	PUNCT
ejpam-1552	141	46	117	117	NUM
ejpam-1552	141	47	362	362	NUM
ejpam-1552	141	48	1772	1772	NUM
ejpam-1552	141	49	205	205	NUM
ejpam-1552	141	50	2456	2456	NUM
ejpam-1552	141	51	worst	bad	ADJ
ejpam-1552	141	52	(	(	PUNCT
ejpam-1552	141	53	4	4	NUM
ejpam-1552	141	54	)	)	PUNCT
ejpam-1552	141	55	36	36	NUM
ejpam-1552	141	56	82	82	NUM
ejpam-1552	141	57	179	179	NUM
ejpam-1552	141	58	492	492	NUM
ejpam-1552	141	59	789	789	NUM
ejpam-1552	141	60	total	total	NOUN
ejpam-1552	141	61	1907	1907	NUM
ejpam-1552	141	62	2222	2222	NUM
ejpam-1552	141	63	2507	2507	NUM
ejpam-1552	141	64	841	841	NUM
ejpam-1552	141	65	7477	7477	NUM
ejpam-1552	141	66	table	table	NOUN
ejpam-1552	141	67	2	2	NUM
ejpam-1552	141	68	gives	give	VERB
ejpam-1552	141	69	the	the	DET
ejpam-1552	141	70	values	value	NOUN
ejpam-1552	141	71	of	of	ADP
ejpam-1552	141	72	the	the	DET
ejpam-1552	141	73	likelihood	likelihood	NOUN
ejpam-1552	141	74	ratio	ratio	NOUN
ejpam-1552	141	75	chi	chi	ADJ
ejpam-1552	141	76	-	-	PUNCT
ejpam-1552	141	77	squared	square	VERB
ejpam-1552	141	78	statistic	statistic	NOUN
ejpam-1552	141	79	g2	g2	PROPN
ejpam-1552	141	80	for	for	ADP
ejpam-1552	141	81	each	each	DET
ejpam-1552	141	82	model	model	NOUN
ejpam-1552	141	83	.	.	PUNCT
ejpam-1552	142	1	the	the	DET
ejpam-1552	142	2	s	s	PROPN
ejpam-1552	142	3	model	model	NOUN
ejpam-1552	142	4	fits	fit	VERB
ejpam-1552	142	5	the	the	DET
ejpam-1552	142	6	data	datum	NOUN
ejpam-1552	142	7	in	in	ADP
ejpam-1552	142	8	table	table	NOUN
ejpam-1552	142	9	1	1	NUM
ejpam-1552	142	10	poorly	poorly	ADV
ejpam-1552	142	11	.	.	PUNCT
ejpam-1552	143	1	we	we	PRON
ejpam-1552	143	2	can	can	AUX
ejpam-1552	143	3	see	see	VERB
ejpam-1552	143	4	from	from	ADP
ejpam-1552	143	5	theorem	theorem	ADJ
ejpam-1552	143	6	1	1	NUM
ejpam-1552	143	7	that	that	SCONJ
ejpam-1552	143	8	the	the	DET
ejpam-1552	143	9	poor	poor	ADJ
ejpam-1552	143	10	fit	fit	NOUN
ejpam-1552	143	11	of	of	ADP
ejpam-1552	143	12	the	the	DET
ejpam-1552	143	13	s	s	PART
ejpam-1552	143	14	model	model	NOUN
ejpam-1552	143	15	is	be	AUX
ejpam-1552	143	16	caused	cause	VERB
ejpam-1552	143	17	by	by	ADP
ejpam-1552	143	18	the	the	DET
ejpam-1552	143	19	influence	influence	NOUN
ejpam-1552	143	20	of	of	ADP
ejpam-1552	143	21	the	the	DET
ejpam-1552	143	22	lack	lack	NOUN
ejpam-1552	143	23	of	of	ADP
ejpam-1552	143	24	structure	structure	NOUN
ejpam-1552	143	25	of	of	ADP
ejpam-1552	143	26	the	the	DET
ejpam-1552	143	27	mh	mh	PROPN
ejpam-1552	143	28	model	model	NOUN
ejpam-1552	143	29	rather	rather	ADV
ejpam-1552	143	30	than	than	ADP
ejpam-1552	143	31	that	that	PRON
ejpam-1552	143	32	of	of	ADP
ejpam-1552	143	33	the	the	DET
ejpam-1552	143	34	qs	qs	PROPN
ejpam-1552	143	35	model	model	NOUN
ejpam-1552	143	36	.	.	PUNCT
ejpam-1552	144	1	table	table	NOUN
ejpam-1552	144	2	2	2	NUM
ejpam-1552	144	3	:	:	PUNCT
ejpam-1552	144	4	likelihood	likelihood	NOUN
ejpam-1552	144	5	ratio	ratio	NOUN
ejpam-1552	144	6	chi	chi	ADJ
ejpam-1552	144	7	-	-	PUNCT
ejpam-1552	144	8	square	square	ADJ
ejpam-1552	144	9	values	value	NOUN
ejpam-1552	144	10	for	for	ADP
ejpam-1552	144	11	models	model	NOUN
ejpam-1552	144	12	applied	apply	VERB
ejpam-1552	144	13	to	to	ADP
ejpam-1552	144	14	the	the	DET
ejpam-1552	144	15	data	datum	NOUN
ejpam-1552	144	16	in	in	ADP
ejpam-1552	144	17	table	table	NOUN
ejpam-1552	144	18	1	1	NUM
ejpam-1552	144	19	.	.	PUNCT
ejpam-1552	145	1	(	(	PUNCT
ejpam-1552	145	2	∗	∗	NOUN
ejpam-1552	145	3	means	mean	VERB
ejpam-1552	145	4	significant	significant	ADJ
ejpam-1552	145	5	at	at	ADP
ejpam-1552	145	6	the	the	DET
ejpam-1552	145	7	0.05	0.05	NUM
ejpam-1552	145	8	level	level	NOUN
ejpam-1552	145	9	.	.	PUNCT
ejpam-1552	145	10	)	)	PUNCT
ejpam-1552	146	1	applied	apply	VERB
ejpam-1552	146	2	degrees	degree	NOUN
ejpam-1552	146	3	of	of	ADP
ejpam-1552	146	4	likelihood	likelihood	NOUN
ejpam-1552	146	5	ratio	ratio	NOUN
ejpam-1552	146	6	models	model	NOUN
ejpam-1552	146	7	freedom	freedom	NOUN
ejpam-1552	146	8	chi	chi	PROPN
ejpam-1552	146	9	-	-	PUNCT
ejpam-1552	146	10	square	square	NOUN
ejpam-1552	146	11	s	s	PART
ejpam-1552	146	12	6	6	NUM
ejpam-1552	146	13	19.25∗	19.25∗	NUM
ejpam-1552	146	14	qs	qs	NOUN
ejpam-1552	146	15	3	3	NUM
ejpam-1552	146	16	7.27	7.27	NUM
ejpam-1552	146	17	mh	mh	NOUN
ejpam-1552	146	18	3	3	NUM
ejpam-1552	146	19	11.99∗	11.99∗	NUM
ejpam-1552	146	20	cs	cs	PROPN
ejpam-1552	146	21	5	5	NUM
ejpam-1552	146	22	7.35	7.35	NUM
ejpam-1552	146	23	ldps	ldps	NOUN
ejpam-1552	146	24	5	5	NUM
ejpam-1552	146	25	7.28	7.28	NUM
ejpam-1552	146	26	2rps	2rps	NUM
ejpam-1552	146	27	4	4	NUM
ejpam-1552	146	28	6.83	6.83	NUM
ejpam-1552	146	29	gs	gs	NOUN
ejpam-1552	146	30	1	1	NUM
ejpam-1552	146	31	11.90∗	11.90∗	NUM
ejpam-1552	146	32	me	i	PRON
ejpam-1552	146	33	1	1	NUM
ejpam-1552	146	34	11.98∗	11.98∗	NUM
ejpam-1552	146	35	cldps	cldps	NOUN
ejpam-1552	146	36	5	5	NUM
ejpam-1552	146	37	8.63	8.63	NUM
ejpam-1552	146	38	c2rps	c2rps	PROPN
ejpam-1552	146	39	4	4	NUM
ejpam-1552	146	40	6.26	6.26	NUM
ejpam-1552	146	41	cqs	cqs	NOUN
ejpam-1552	146	42	3	3	NUM
ejpam-1552	146	43	8.43∗	8.43∗	NUM
ejpam-1552	146	44	we	we	PRON
ejpam-1552	146	45	can	can	AUX
ejpam-1552	146	46	also	also	ADV
ejpam-1552	146	47	see	see	VERB
ejpam-1552	146	48	from	from	ADP
ejpam-1552	146	49	theorem	theorem	ADJ
ejpam-1552	146	50	2	2	NUM
ejpam-1552	146	51	(	(	PUNCT
ejpam-1552	146	52	theorem	theorem	NOUN
ejpam-1552	146	53	3	3	NUM
ejpam-1552	146	54	)	)	PUNCT
ejpam-1552	146	55	that	that	SCONJ
ejpam-1552	146	56	the	the	DET
ejpam-1552	146	57	poor	poor	ADJ
ejpam-1552	146	58	fit	fit	NOUN
ejpam-1552	146	59	of	of	ADP
ejpam-1552	146	60	the	the	DET
ejpam-1552	146	61	s	s	PART
ejpam-1552	146	62	model	model	NOUN
ejpam-1552	146	63	is	be	AUX
ejpam-1552	146	64	caused	cause	VERB
ejpam-1552	146	65	by	by	ADP
ejpam-1552	146	66	the	the	DET
ejpam-1552	146	67	influence	influence	NOUN
ejpam-1552	146	68	of	of	ADP
ejpam-1552	146	69	the	the	DET
ejpam-1552	146	70	lack	lack	NOUN
ejpam-1552	146	71	of	of	ADP
ejpam-1552	146	72	structure	structure	NOUN
ejpam-1552	146	73	of	of	ADP
ejpam-1552	146	74	the	the	DET
ejpam-1552	146	75	me	me	PROPN
ejpam-1552	146	76	model	model	NOUN
ejpam-1552	146	77	(	(	PUNCT
ejpam-1552	146	78	the	the	DET
ejpam-1552	146	79	gs	gs	NOUN
ejpam-1552	146	80	and	and	CCONJ
ejpam-1552	146	81	me	i	PRON
ejpam-1552	146	82	models	model	NOUN
ejpam-1552	146	83	)	)	PUNCT
ejpam-1552	146	84	rather	rather	ADV
ejpam-1552	146	85	than	than	ADP
ejpam-1552	146	86	that	that	PRON
ejpam-1552	146	87	of	of	ADP
ejpam-1552	146	88	the	the	DET
ejpam-1552	146	89	ldps	ldps	NOUN
ejpam-1552	146	90	(	(	PUNCT
ejpam-1552	146	91	the	the	DET
ejpam-1552	146	92	2rps	2rps	NOUN
ejpam-1552	146	93	)	)	PUNCT
ejpam-1552	146	94	model	model	NOUN
ejpam-1552	146	95	.	.	PUNCT
ejpam-1552	147	1	references	reference	NOUN
ejpam-1552	147	2	305	305	NUM
ejpam-1552	147	3	the	the	DET
ejpam-1552	147	4	s	s	NOUN
ejpam-1552	147	5	model	model	NOUN
ejpam-1552	147	6	also	also	ADV
ejpam-1552	147	7	indicates	indicate	VERB
ejpam-1552	147	8	the	the	DET
ejpam-1552	147	9	structure	structure	NOUN
ejpam-1552	147	10	of	of	ADP
ejpam-1552	147	11	symmetry	symmetry	NOUN
ejpam-1552	147	12	of	of	ADP
ejpam-1552	147	13	cumulative	cumulative	ADJ
ejpam-1552	147	14	probabilities	probability	NOUN
ejpam-1552	147	15	{	{	PUNCT
ejpam-1552	147	16	gi	gi	NOUN
ejpam-1552	147	17	j	j	NOUN
ejpam-1552	147	18	}	}	PUNCT
ejpam-1552	147	19	,	,	PUNCT
ejpam-1552	147	20	i	i	PROPN
ejpam-1552	147	21	6=	6=	PROPN
ejpam-1552	147	22	j	j	PROPN
ejpam-1552	147	23	,	,	PUNCT
ejpam-1552	147	24	instead	instead	ADV
ejpam-1552	147	25	of	of	ADP
ejpam-1552	147	26	the	the	DET
ejpam-1552	147	27	cell	cell	NOUN
ejpam-1552	147	28	probabilities	probability	NOUN
ejpam-1552	147	29	{	{	PUNCT
ejpam-1552	147	30	pi	pi	NOUN
ejpam-1552	147	31	j	j	PROPN
ejpam-1552	147	32	}	}	PUNCT
ejpam-1552	147	33	,	,	PUNCT
ejpam-1552	147	34	i	i	PROPN
ejpam-1552	147	35	6=	6=	PROPN
ejpam-1552	147	36	j.	j.	PROPN
ejpam-1552	147	37	therefore	therefore	ADV
ejpam-1552	147	38	we	we	PRON
ejpam-1552	147	39	shall	shall	AUX
ejpam-1552	147	40	next	next	ADV
ejpam-1552	147	41	consider	consider	VERB
ejpam-1552	147	42	the	the	DET
ejpam-1552	147	43	reason	reason	NOUN
ejpam-1552	147	44	why	why	SCONJ
ejpam-1552	147	45	the	the	DET
ejpam-1552	147	46	s	s	NOUN
ejpam-1552	147	47	model	model	NOUN
ejpam-1552	147	48	fits	fit	VERB
ejpam-1552	147	49	the	the	DET
ejpam-1552	147	50	data	datum	NOUN
ejpam-1552	147	51	in	in	ADP
ejpam-1552	147	52	table	table	NOUN
ejpam-1552	147	53	1	1	NUM
ejpam-1552	147	54	poorly	poorly	ADV
ejpam-1552	147	55	using	use	VERB
ejpam-1552	147	56	the	the	DET
ejpam-1552	147	57	models	model	NOUN
ejpam-1552	147	58	which	which	PRON
ejpam-1552	147	59	describe	describe	VERB
ejpam-1552	147	60	the	the	DET
ejpam-1552	147	61	structure	structure	NOUN
ejpam-1552	147	62	of	of	ADP
ejpam-1552	147	63	cumulative	cumulative	ADJ
ejpam-1552	147	64	probabilities	probability	NOUN
ejpam-1552	147	65	.	.	PUNCT
ejpam-1552	148	1	we	we	PRON
ejpam-1552	148	2	see	see	VERB
ejpam-1552	148	3	from	from	ADP
ejpam-1552	148	4	theorem	theorem	ADJ
ejpam-1552	148	5	4	4	NUM
ejpam-1552	148	6	that	that	SCONJ
ejpam-1552	148	7	the	the	DET
ejpam-1552	148	8	poor	poor	ADJ
ejpam-1552	148	9	fit	fit	NOUN
ejpam-1552	148	10	of	of	ADP
ejpam-1552	148	11	the	the	DET
ejpam-1552	148	12	s	s	PART
ejpam-1552	148	13	model	model	NOUN
ejpam-1552	148	14	is	be	AUX
ejpam-1552	148	15	caused	cause	VERB
ejpam-1552	148	16	by	by	ADP
ejpam-1552	148	17	the	the	DET
ejpam-1552	148	18	influence	influence	NOUN
ejpam-1552	148	19	of	of	ADP
ejpam-1552	148	20	the	the	DET
ejpam-1552	148	21	lack	lack	NOUN
ejpam-1552	148	22	of	of	ADP
ejpam-1552	148	23	structures	structure	NOUN
ejpam-1552	148	24	of	of	ADP
ejpam-1552	148	25	both	both	CCONJ
ejpam-1552	148	26	the	the	DET
ejpam-1552	148	27	cqs	cqs	NOUN
ejpam-1552	148	28	and	and	CCONJ
ejpam-1552	148	29	mh	mh	PROPN
ejpam-1552	148	30	models	model	NOUN
ejpam-1552	148	31	.	.	PUNCT
ejpam-1552	149	1	we	we	PRON
ejpam-1552	149	2	also	also	ADV
ejpam-1552	149	3	see	see	VERB
ejpam-1552	149	4	from	from	ADP
ejpam-1552	149	5	theorem	theorem	ADJ
ejpam-1552	149	6	5	5	NUM
ejpam-1552	149	7	that	that	SCONJ
ejpam-1552	149	8	the	the	DET
ejpam-1552	149	9	poor	poor	ADJ
ejpam-1552	149	10	fit	fit	NOUN
ejpam-1552	149	11	of	of	ADP
ejpam-1552	149	12	the	the	DET
ejpam-1552	149	13	s	s	PART
ejpam-1552	149	14	model	model	NOUN
ejpam-1552	149	15	is	be	AUX
ejpam-1552	149	16	caused	cause	VERB
ejpam-1552	149	17	by	by	ADP
ejpam-1552	149	18	the	the	DET
ejpam-1552	149	19	influence	influence	NOUN
ejpam-1552	149	20	of	of	ADP
ejpam-1552	149	21	the	the	DET
ejpam-1552	149	22	lack	lack	NOUN
ejpam-1552	149	23	of	of	ADP
ejpam-1552	149	24	structure	structure	NOUN
ejpam-1552	149	25	of	of	ADP
ejpam-1552	149	26	the	the	DET
ejpam-1552	149	27	me	me	PROPN
ejpam-1552	149	28	model	model	NOUN
ejpam-1552	149	29	rather	rather	ADV
ejpam-1552	149	30	than	than	ADP
ejpam-1552	149	31	the	the	DET
ejpam-1552	149	32	cldps	cldps	NOUN
ejpam-1552	149	33	model	model	NOUN
ejpam-1552	149	34	.	.	PUNCT
ejpam-1552	150	1	in	in	ADP
ejpam-1552	150	2	more	more	ADJ
ejpam-1552	150	3	details	detail	NOUN
ejpam-1552	150	4	,	,	PUNCT
ejpam-1552	150	5	we	we	PRON
ejpam-1552	150	6	can	can	AUX
ejpam-1552	150	7	see	see	VERB
ejpam-1552	150	8	from	from	ADP
ejpam-1552	150	9	theorem	theorem	NOUN
ejpam-1552	150	10	6	6	NUM
ejpam-1552	150	11	that	that	SCONJ
ejpam-1552	150	12	the	the	DET
ejpam-1552	150	13	poor	poor	ADJ
ejpam-1552	150	14	fit	fit	NOUN
ejpam-1552	150	15	of	of	ADP
ejpam-1552	150	16	the	the	DET
ejpam-1552	150	17	s	s	PART
ejpam-1552	150	18	model	model	NOUN
ejpam-1552	150	19	is	be	AUX
ejpam-1552	150	20	caused	cause	VERB
ejpam-1552	150	21	by	by	ADP
ejpam-1552	150	22	the	the	DET
ejpam-1552	150	23	influence	influence	NOUN
ejpam-1552	150	24	of	of	ADP
ejpam-1552	150	25	the	the	DET
ejpam-1552	150	26	lack	lack	NOUN
ejpam-1552	150	27	of	of	ADP
ejpam-1552	150	28	structures	structure	NOUN
ejpam-1552	150	29	of	of	ADP
ejpam-1552	150	30	the	the	DET
ejpam-1552	150	31	gs	gs	NOUN
ejpam-1552	150	32	and	and	CCONJ
ejpam-1552	150	33	me	i	PRON
ejpam-1552	150	34	models	model	NOUN
ejpam-1552	150	35	rather	rather	ADV
ejpam-1552	150	36	than	than	ADP
ejpam-1552	150	37	the	the	DET
ejpam-1552	150	38	c2rps	c2rps	PROPN
ejpam-1552	150	39	model	model	NOUN
ejpam-1552	150	40	.	.	PUNCT
ejpam-1552	151	1	5	5	X
ejpam-1552	151	2	.	.	X
ejpam-1552	151	3	concluding	conclude	VERB
ejpam-1552	151	4	remarks	remark	NOUN
ejpam-1552	151	5	theorems	theorem	NOUN
ejpam-1552	151	6	1	1	NUM
ejpam-1552	151	7	through	through	ADP
ejpam-1552	151	8	6	6	NUM
ejpam-1552	151	9	would	would	AUX
ejpam-1552	151	10	be	be	AUX
ejpam-1552	151	11	useful	useful	ADJ
ejpam-1552	151	12	for	for	ADP
ejpam-1552	151	13	seeing	see	VERB
ejpam-1552	151	14	the	the	DET
ejpam-1552	151	15	reason	reason	NOUN
ejpam-1552	151	16	for	for	ADP
ejpam-1552	151	17	poor	poor	ADJ
ejpam-1552	151	18	fit	fit	NOUN
ejpam-1552	151	19	of	of	ADP
ejpam-1552	151	20	the	the	DET
ejpam-1552	151	21	s	s	NOUN
ejpam-1552	151	22	model	model	NOUN
ejpam-1552	151	23	when	when	SCONJ
ejpam-1552	151	24	the	the	DET
ejpam-1552	151	25	s	s	PROPN
ejpam-1552	151	26	model	model	NOUN
ejpam-1552	151	27	fits	fit	VERB
ejpam-1552	151	28	the	the	DET
ejpam-1552	151	29	data	datum	NOUN
ejpam-1552	151	30	poorly	poorly	ADV
ejpam-1552	151	31	.	.	PUNCT
ejpam-1552	152	1	when	when	SCONJ
ejpam-1552	152	2	we	we	PRON
ejpam-1552	152	3	are	be	AUX
ejpam-1552	152	4	interested	interested	ADJ
ejpam-1552	152	5	in	in	ADP
ejpam-1552	152	6	the	the	DET
ejpam-1552	152	7	structure	structure	NOUN
ejpam-1552	152	8	of	of	ADP
ejpam-1552	152	9	symmetry	symmetry	NOUN
ejpam-1552	152	10	of	of	ADP
ejpam-1552	152	11	cumulative	cumulative	ADJ
ejpam-1552	152	12	probabilities	probability	NOUN
ejpam-1552	152	13	{	{	PUNCT
ejpam-1552	152	14	gi	gi	NOUN
ejpam-1552	152	15	j	j	NOUN
ejpam-1552	152	16	}	}	PUNCT
ejpam-1552	152	17	,	,	PUNCT
ejpam-1552	152	18	i	i	PROPN
ejpam-1552	152	19	6=	6=	PROPN
ejpam-1552	152	20	j	j	PROPN
ejpam-1552	152	21	,	,	PUNCT
ejpam-1552	152	22	instead	instead	ADV
ejpam-1552	152	23	of	of	ADP
ejpam-1552	152	24	the	the	DET
ejpam-1552	152	25	cell	cell	NOUN
ejpam-1552	152	26	probabilities	probability	NOUN
ejpam-1552	152	27	{	{	PUNCT
ejpam-1552	152	28	pi	pi	NOUN
ejpam-1552	152	29	j	j	PROPN
ejpam-1552	152	30	}	}	PUNCT
ejpam-1552	152	31	,	,	PUNCT
ejpam-1552	152	32	i	i	PROPN
ejpam-1552	152	33	6=	6=	PROPN
ejpam-1552	152	34	j	j	PROPN
ejpam-1552	152	35	,	,	PUNCT
ejpam-1552	152	36	theorems	theorem	VERB
ejpam-1552	152	37	4	4	NUM
ejpam-1552	152	38	,	,	PUNCT
ejpam-1552	152	39	5	5	NUM
ejpam-1552	152	40	and	and	CCONJ
ejpam-1552	152	41	6	6	NUM
ejpam-1552	152	42	would	would	AUX
ejpam-1552	152	43	be	be	AUX
ejpam-1552	152	44	useful	useful	ADJ
ejpam-1552	152	45	.	.	PUNCT
ejpam-1552	153	1	especially	especially	ADV
ejpam-1552	153	2	,	,	PUNCT
ejpam-1552	153	3	theorem	theorem	VERB
ejpam-1552	153	4	6	6	NUM
ejpam-1552	153	5	rather	rather	ADV
ejpam-1552	153	6	than	than	ADP
ejpam-1552	153	7	theorem	theorem	ADJ
ejpam-1552	153	8	5	5	NUM
ejpam-1552	153	9	would	would	AUX
ejpam-1552	153	10	be	be	AUX
ejpam-1552	153	11	useful	useful	ADJ
ejpam-1552	153	12	for	for	ADP
ejpam-1552	153	13	seeing	see	VERB
ejpam-1552	153	14	in	in	ADP
ejpam-1552	153	15	more	more	ADJ
ejpam-1552	153	16	details	detail	NOUN
ejpam-1552	153	17	the	the	DET
ejpam-1552	153	18	reason	reason	NOUN
ejpam-1552	153	19	for	for	ADP
ejpam-1552	153	20	poor	poor	ADJ
ejpam-1552	153	21	fit	fit	NOUN
ejpam-1552	153	22	of	of	ADP
ejpam-1552	153	23	the	the	DET
ejpam-1552	153	24	s	s	NOUN
ejpam-1552	153	25	model	model	NOUN
ejpam-1552	153	26	when	when	SCONJ
ejpam-1552	153	27	the	the	DET
ejpam-1552	153	28	s	s	PROPN
ejpam-1552	153	29	model	model	NOUN
ejpam-1552	153	30	fits	fit	VERB
ejpam-1552	153	31	the	the	DET
ejpam-1552	153	32	data	datum	NOUN
ejpam-1552	153	33	poorly	poorly	ADV
ejpam-1552	153	34	.	.	PUNCT
ejpam-1552	154	1	references	reference	NOUN
ejpam-1552	154	2	[	[	X
ejpam-1552	154	3	1	1	X
ejpam-1552	154	4	]	]	PUNCT
ejpam-1552	154	5	a	a	DET
ejpam-1552	154	6	agresti	agresti	NOUN
ejpam-1552	154	7	.	.	PUNCT
ejpam-1552	155	1	a	a	DET
ejpam-1552	155	2	simple	simple	ADJ
ejpam-1552	155	3	diagonals	diagonal	NOUN
ejpam-1552	155	4	-	-	PUNCT
ejpam-1552	155	5	parameter	parameter	NOUN
ejpam-1552	155	6	symmetry	symmetry	NOUN
ejpam-1552	155	7	and	and	CCONJ
ejpam-1552	155	8	quasi	quasi	ADJ
ejpam-1552	155	9	-	-	ADJ
ejpam-1552	155	10	symmetry	symmetry	ADJ
ejpam-1552	155	11	model	model	NOUN
ejpam-1552	155	12	.	.	PUNCT
ejpam-1552	156	1	statistics	statistic	NOUN
ejpam-1552	156	2	and	and	CCONJ
ejpam-1552	156	3	probability	probability	NOUN
ejpam-1552	156	4	letters	letter	NOUN
ejpam-1552	156	5	,	,	PUNCT
ejpam-1552	156	6	1:313–316	1:313–316	NUM
ejpam-1552	156	7	,	,	PUNCT
ejpam-1552	156	8	1983	1983	NUM
ejpam-1552	156	9	.	.	PUNCT
ejpam-1552	157	1	[	[	X
ejpam-1552	157	2	2	2	NUM
ejpam-1552	157	3	]	]	X
ejpam-1552	157	4	y	y	PROPN
ejpam-1552	157	5	m	m	NOUN
ejpam-1552	157	6	m	m	VERB
ejpam-1552	157	7	bishop	bishop	NOUN
ejpam-1552	157	8	,	,	PUNCT
ejpam-1552	157	9	s	s	PROPN
ejpam-1552	157	10	e	e	PROPN
ejpam-1552	157	11	fienberg	fienberg	PROPN
ejpam-1552	157	12	,	,	PUNCT
ejpam-1552	157	13	and	and	CCONJ
ejpam-1552	157	14	p	p	PROPN
ejpam-1552	157	15	w	w	PROPN
ejpam-1552	157	16	holland	holland	PROPN
ejpam-1552	157	17	.	.	PUNCT
ejpam-1552	158	1	discrete	discrete	ADJ
ejpam-1552	158	2	multivariate	multivariate	NOUN
ejpam-1552	158	3	analysis	analysis	NOUN
ejpam-1552	158	4	:	:	PUNCT
ejpam-1552	158	5	theory	theory	NOUN
ejpam-1552	158	6	and	and	CCONJ
ejpam-1552	158	7	practice	practice	NOUN
ejpam-1552	158	8	.	.	PUNCT
ejpam-1552	159	1	the	the	DET
ejpam-1552	159	2	mit	mit	PROPN
ejpam-1552	159	3	press	press	PROPN
ejpam-1552	159	4	,	,	PUNCT
ejpam-1552	159	5	cambridge	cambridge	PROPN
ejpam-1552	159	6	,	,	PUNCT
ejpam-1552	159	7	massachusetts	massachusetts	PROPN
ejpam-1552	159	8	,	,	PUNCT
ejpam-1552	159	9	1975	1975	NUM
ejpam-1552	159	10	.	.	PUNCT
ejpam-1552	160	1	[	[	X
ejpam-1552	160	2	3	3	X
ejpam-1552	160	3	]	]	PUNCT
ejpam-1552	160	4	a	a	DET
ejpam-1552	160	5	h	h	NOUN
ejpam-1552	160	6	bowker	bowker	NOUN
ejpam-1552	160	7	.	.	PUNCT
ejpam-1552	161	1	a	a	DET
ejpam-1552	161	2	test	test	NOUN
ejpam-1552	161	3	for	for	ADP
ejpam-1552	161	4	symmetry	symmetry	NOUN
ejpam-1552	161	5	in	in	ADP
ejpam-1552	161	6	contingency	contingency	NOUN
ejpam-1552	161	7	tables	table	NOUN
ejpam-1552	161	8	.	.	PUNCT
ejpam-1552	162	1	journal	journal	NOUN
ejpam-1552	162	2	of	of	ADP
ejpam-1552	162	3	the	the	DET
ejpam-1552	162	4	american	american	PROPN
ejpam-1552	162	5	statistical	statistical	PROPN
ejpam-1552	162	6	association	association	PROPN
ejpam-1552	162	7	,	,	PUNCT
ejpam-1552	162	8	43:572–574	43:572–574	PROPN
ejpam-1552	162	9	,	,	PUNCT
ejpam-1552	162	10	1948	1948	NUM
ejpam-1552	162	11	.	.	PUNCT
ejpam-1552	163	1	[	[	X
ejpam-1552	163	2	4	4	X
ejpam-1552	163	3	]	]	X
ejpam-1552	163	4	h	h	NOUN
ejpam-1552	163	5	caussinus	caussinus	NOUN
ejpam-1552	163	6	.	.	PUNCT
ejpam-1552	164	1	contribution	contribution	NOUN
ejpam-1552	164	2	à	à	X
ejpam-1552	164	3	l’analyse	l’analyse	NOUN
ejpam-1552	164	4	statistique	statistique	PROPN
ejpam-1552	164	5	des	des	PROPN
ejpam-1552	164	6	tableaux	tableaux	X
ejpam-1552	164	7	de	de	PROPN
ejpam-1552	164	8	corrélation	corrélation	NOUN
ejpam-1552	164	9	.	.	PUNCT
ejpam-1552	165	1	annales	annales	PROPN
ejpam-1552	165	2	de	de	PROPN
ejpam-1552	165	3	la	la	X
ejpam-1552	165	4	faculté	faculté	PROPN
ejpam-1552	165	5	des	des	PROPN
ejpam-1552	165	6	sciences	sciences	PROPN
ejpam-1552	165	7	de	de	X
ejpam-1552	165	8	l’université	l’université	X
ejpam-1552	165	9	de	de	X
ejpam-1552	165	10	toulouse	toulouse	NOUN
ejpam-1552	165	11	,	,	PUNCT
ejpam-1552	165	12	29:77–182	29:77–182	NUM
ejpam-1552	165	13	,	,	PUNCT
ejpam-1552	165	14	1965	1965	NUM
ejpam-1552	165	15	.	.	PUNCT
ejpam-1552	166	1	[	[	X
ejpam-1552	166	2	5	5	NUM
ejpam-1552	166	3	]	]	PUNCT
ejpam-1552	166	4	l	l	NOUN
ejpam-1552	166	5	a	a	DET
ejpam-1552	166	6	goodman	goodman	PROPN
ejpam-1552	166	7	.	.	PUNCT
ejpam-1552	167	1	multiplicative	multiplicative	ADJ
ejpam-1552	167	2	models	model	NOUN
ejpam-1552	167	3	for	for	ADP
ejpam-1552	167	4	square	square	ADJ
ejpam-1552	167	5	contingency	contingency	NOUN
ejpam-1552	167	6	tables	table	NOUN
ejpam-1552	167	7	with	with	ADP
ejpam-1552	167	8	ordered	order	VERB
ejpam-1552	167	9	categories	category	NOUN
ejpam-1552	167	10	.	.	PUNCT
ejpam-1552	168	1	biometrika	biometrika	PROPN
ejpam-1552	168	2	,	,	PUNCT
ejpam-1552	168	3	66:413–418	66:413–418	PROPN
ejpam-1552	168	4	,	,	PUNCT
ejpam-1552	168	5	1979	1979	NUM
ejpam-1552	168	6	.	.	PUNCT
ejpam-1552	169	1	[	[	X
ejpam-1552	169	2	6	6	NUM
ejpam-1552	169	3	]	]	PUNCT
ejpam-1552	169	4	p	p	X
ejpam-1552	169	5	mccullagh	mccullagh	NOUN
ejpam-1552	169	6	.	.	PUNCT
ejpam-1552	170	1	a	a	DET
ejpam-1552	170	2	class	class	NOUN
ejpam-1552	170	3	of	of	ADP
ejpam-1552	170	4	parametric	parametric	ADJ
ejpam-1552	170	5	models	model	NOUN
ejpam-1552	170	6	for	for	ADP
ejpam-1552	170	7	the	the	DET
ejpam-1552	170	8	analysis	analysis	NOUN
ejpam-1552	170	9	of	of	ADP
ejpam-1552	170	10	square	square	ADJ
ejpam-1552	170	11	contingency	contingency	NOUN
ejpam-1552	170	12	tables	table	NOUN
ejpam-1552	170	13	with	with	ADP
ejpam-1552	170	14	ordered	order	VERB
ejpam-1552	170	15	categories	category	NOUN
ejpam-1552	170	16	.	.	PUNCT
ejpam-1552	171	1	biometrika	biometrika	NOUN
ejpam-1552	171	2	,	,	PUNCT
ejpam-1552	171	3	65:413–418	65:413–418	PROPN
ejpam-1552	171	4	,	,	PUNCT
ejpam-1552	171	5	1978	1978	NUM
ejpam-1552	171	6	.	.	PUNCT
ejpam-1552	172	1	[	[	X
ejpam-1552	172	2	7	7	NUM
ejpam-1552	172	3	]	]	PUNCT
ejpam-1552	172	4	n	n	DET
ejpam-1552	172	5	miyamoto	miyamoto	NOUN
ejpam-1552	172	6	,	,	PUNCT
ejpam-1552	172	7	w	w	PROPN
ejpam-1552	172	8	ohtsuka	ohtsuka	PROPN
ejpam-1552	172	9	,	,	PUNCT
ejpam-1552	172	10	and	and	CCONJ
ejpam-1552	172	11	s	s	VERB
ejpam-1552	172	12	tomizawa	tomizawa	NOUN
ejpam-1552	172	13	.	.	PUNCT
ejpam-1552	173	1	linear	linear	ADJ
ejpam-1552	173	2	diagonals	diagonal	NOUN
ejpam-1552	173	3	-	-	PUNCT
ejpam-1552	173	4	parameter	parameter	NOUN
ejpam-1552	173	5	symmetry	symmetry	NOUN
ejpam-1552	173	6	and	and	CCONJ
ejpam-1552	173	7	quasi	quasi	ADJ
ejpam-1552	173	8	-	-	NOUN
ejpam-1552	173	9	symmetry	symmetry	ADJ
ejpam-1552	173	10	models	model	NOUN
ejpam-1552	173	11	for	for	ADP
ejpam-1552	173	12	cumulative	cumulative	ADJ
ejpam-1552	173	13	probabilities	probability	NOUN
ejpam-1552	173	14	in	in	ADP
ejpam-1552	173	15	square	square	ADJ
ejpam-1552	173	16	contingency	contingency	NOUN
ejpam-1552	173	17	tables	table	NOUN
ejpam-1552	173	18	with	with	ADP
ejpam-1552	173	19	ordered	order	VERB
ejpam-1552	173	20	categories	category	NOUN
ejpam-1552	173	21	.	.	PUNCT
ejpam-1552	174	1	biometrical	biometrical	ADJ
ejpam-1552	174	2	journal	journal	PROPN
ejpam-1552	174	3	,	,	PUNCT
ejpam-1552	174	4	46:664–674	46:664–674	PROPN
ejpam-1552	174	5	,	,	PUNCT
ejpam-1552	174	6	2004	2004	NUM
ejpam-1552	174	7	.	.	PUNCT
ejpam-1552	175	1	[	[	X
ejpam-1552	175	2	8	8	NUM
ejpam-1552	175	3	]	]	PUNCT
ejpam-1552	175	4	a	a	DET
ejpam-1552	175	5	stuart	stuart	PROPN
ejpam-1552	175	6	.	.	PUNCT
ejpam-1552	176	1	a	a	DET
ejpam-1552	176	2	test	test	NOUN
ejpam-1552	176	3	for	for	ADP
ejpam-1552	176	4	homogeneity	homogeneity	NOUN
ejpam-1552	176	5	of	of	ADP
ejpam-1552	176	6	the	the	DET
ejpam-1552	176	7	marginal	marginal	ADJ
ejpam-1552	176	8	distributions	distribution	NOUN
ejpam-1552	176	9	in	in	ADP
ejpam-1552	176	10	a	a	DET
ejpam-1552	176	11	two	two	NUM
ejpam-1552	176	12	-	-	PUNCT
ejpam-1552	176	13	way	way	NOUN
ejpam-1552	176	14	classification	classification	NOUN
ejpam-1552	176	15	.	.	PUNCT
ejpam-1552	177	1	biometrika	biometrika	NOUN
ejpam-1552	177	2	,	,	PUNCT
ejpam-1552	177	3	42:412–416	42:412–416	PROPN
ejpam-1552	177	4	,	,	PUNCT
ejpam-1552	177	5	1955	1955	NUM
ejpam-1552	177	6	.	.	PUNCT
ejpam-1552	178	1	references	reference	NOUN
ejpam-1552	178	2	306	306	NUM
ejpam-1552	179	1	[	[	X
ejpam-1552	179	2	9	9	NUM
ejpam-1552	179	3	]	]	X
ejpam-1552	179	4	k	k	PROPN
ejpam-1552	179	5	tahata	tahata	PROPN
ejpam-1552	179	6	and	and	CCONJ
ejpam-1552	179	7	s	s	NOUN
ejpam-1552	179	8	tomizawa	tomizawa	NOUN
ejpam-1552	179	9	.	.	PUNCT
ejpam-1552	180	1	decomposition	decomposition	NOUN
ejpam-1552	180	2	of	of	ADP
ejpam-1552	180	3	symmetry	symmetry	NOUN
ejpam-1552	180	4	using	use	VERB
ejpam-1552	180	5	two	two	NUM
ejpam-1552	180	6	-	-	PUNCT
ejpam-1552	180	7	ratios	ratio	NOUN
ejpam-1552	180	8	-	-	PUNCT
ejpam-1552	180	9	parameter	parameter	NOUN
ejpam-1552	180	10	symmetry	symmetry	NOUN
ejpam-1552	180	11	model	model	NOUN
ejpam-1552	180	12	and	and	CCONJ
ejpam-1552	180	13	orthogonality	orthogonality	NOUN
ejpam-1552	180	14	for	for	ADP
ejpam-1552	180	15	square	square	ADJ
ejpam-1552	180	16	contingency	contingency	NOUN
ejpam-1552	180	17	tables	table	NOUN
ejpam-1552	180	18	.	.	PUNCT
ejpam-1552	181	1	journal	journal	PROPN
ejpam-1552	181	2	of	of	ADP
ejpam-1552	181	3	statistics	statistic	NOUN
ejpam-1552	181	4	:	:	PUNCT
ejpam-1552	181	5	advances	advance	NOUN
ejpam-1552	181	6	in	in	ADP
ejpam-1552	181	7	theory	theory	NOUN
ejpam-1552	181	8	and	and	CCONJ
ejpam-1552	181	9	applications	application	NOUN
ejpam-1552	181	10	,	,	PUNCT
ejpam-1552	181	11	1:19–33	1:19–33	NUM
ejpam-1552	181	12	,	,	PUNCT
ejpam-1552	181	13	2009	2009	NUM
ejpam-1552	181	14	.	.	PUNCT
ejpam-1552	182	1	[	[	X
ejpam-1552	182	2	10	10	NUM
ejpam-1552	182	3	]	]	X
ejpam-1552	182	4	k	k	PROPN
ejpam-1552	182	5	tahata	tahata	PROPN
ejpam-1552	182	6	,	,	PUNCT
ejpam-1552	182	7	h	h	NOUN
ejpam-1552	182	8	yamamoto	yamamoto	NOUN
ejpam-1552	182	9	,	,	PUNCT
ejpam-1552	182	10	and	and	CCONJ
ejpam-1552	182	11	s	s	VERB
ejpam-1552	182	12	tomizawa	tomizawa	NOUN
ejpam-1552	182	13	.	.	PUNCT
ejpam-1552	183	1	orthogonality	orthogonality	NOUN
ejpam-1552	183	2	of	of	ADP
ejpam-1552	183	3	decompositions	decomposition	NOUN
ejpam-1552	183	4	of	of	ADP
ejpam-1552	183	5	symmetry	symmetry	NOUN
ejpam-1552	183	6	into	into	ADP
ejpam-1552	183	7	extended	extended	ADJ
ejpam-1552	183	8	symmetry	symmetry	NOUN
ejpam-1552	183	9	and	and	CCONJ
ejpam-1552	183	10	marginal	marginal	ADJ
ejpam-1552	183	11	equimoment	equimoment	NOUN
ejpam-1552	183	12	for	for	ADP
ejpam-1552	183	13	multi	multi	ADJ
ejpam-1552	183	14	-	-	ADJ
ejpam-1552	183	15	way	way	ADJ
ejpam-1552	183	16	tables	table	NOUN
ejpam-1552	183	17	with	with	ADP
ejpam-1552	183	18	ordered	order	VERB
ejpam-1552	183	19	categories	category	NOUN
ejpam-1552	183	20	.	.	PUNCT
ejpam-1552	184	1	austrian	austrian	ADJ
ejpam-1552	184	2	journal	journal	PROPN
ejpam-1552	184	3	of	of	ADP
ejpam-1552	184	4	statistics	statistic	NOUN
ejpam-1552	184	5	,	,	PUNCT
ejpam-1552	184	6	37:185–194	37:185–194	PROPN
ejpam-1552	184	7	,	,	PUNCT
ejpam-1552	184	8	2008	2008	NUM
ejpam-1552	184	9	.	.	PUNCT
ejpam-1552	185	1	[	[	X
ejpam-1552	185	2	11	11	NUM
ejpam-1552	185	3	]	]	X
ejpam-1552	185	4	s	s	VERB
ejpam-1552	185	5	tomizawa	tomizawa	NOUN
ejpam-1552	185	6	.	.	PUNCT
ejpam-1552	186	1	decompositions	decomposition	NOUN
ejpam-1552	186	2	for	for	ADP
ejpam-1552	186	3	2	2	NUM
ejpam-1552	186	4	-	-	PUNCT
ejpam-1552	186	5	ratios	ratio	NOUN
ejpam-1552	186	6	-	-	PUNCT
ejpam-1552	186	7	parameter	parameter	NOUN
ejpam-1552	186	8	symmetry	symmetry	NOUN
ejpam-1552	186	9	model	model	NOUN
ejpam-1552	186	10	in	in	ADP
ejpam-1552	186	11	square	square	ADJ
ejpam-1552	186	12	contingency	contingency	NOUN
ejpam-1552	186	13	tables	table	NOUN
ejpam-1552	186	14	with	with	ADP
ejpam-1552	186	15	ordered	order	VERB
ejpam-1552	186	16	categories	category	NOUN
ejpam-1552	186	17	.	.	PUNCT
ejpam-1552	187	1	biometrical	biometrical	ADJ
ejpam-1552	187	2	journal	journal	PROPN
ejpam-1552	187	3	,	,	PUNCT
ejpam-1552	187	4	29:45–55	29:45–55	PROPN
ejpam-1552	187	5	,	,	PUNCT
ejpam-1552	187	6	1987	1987	NUM
ejpam-1552	187	7	.	.	PUNCT
ejpam-1552	188	1	[	[	X
ejpam-1552	188	2	12	12	NUM
ejpam-1552	188	3	]	]	X
ejpam-1552	188	4	s	s	VERB
ejpam-1552	188	5	tomizawa	tomizawa	PROPN
ejpam-1552	188	6	.	.	PUNCT
ejpam-1552	189	1	diagonals	diagonal	NOUN
ejpam-1552	189	2	-	-	PUNCT
ejpam-1552	189	3	parameter	parameter	NOUN
ejpam-1552	189	4	symmetry	symmetry	NOUN
ejpam-1552	189	5	model	model	NOUN
ejpam-1552	189	6	for	for	ADP
ejpam-1552	189	7	cumulative	cumulative	ADJ
ejpam-1552	189	8	probabilities	probability	NOUN
ejpam-1552	189	9	in	in	ADP
ejpam-1552	189	10	square	square	ADJ
ejpam-1552	189	11	contingency	contingency	NOUN
ejpam-1552	189	12	tables	table	NOUN
ejpam-1552	189	13	with	with	ADP
ejpam-1552	189	14	ordered	order	VERB
ejpam-1552	189	15	categories	category	NOUN
ejpam-1552	189	16	.	.	PUNCT
ejpam-1552	190	1	biometrics	biometric	NOUN
ejpam-1552	190	2	,	,	PUNCT
ejpam-1552	190	3	49:883–887	49:883–887	PROPN
ejpam-1552	190	4	,	,	PUNCT
ejpam-1552	190	5	1993	1993	NUM
ejpam-1552	190	6	.	.	PUNCT
ejpam-1552	191	1	[	[	X
ejpam-1552	191	2	13	13	NUM
ejpam-1552	191	3	]	]	SYM
ejpam-1552	191	4	s	s	PART
ejpam-1552	191	5	tomizawa	tomizawa	PROPN
ejpam-1552	191	6	,	,	PUNCT
ejpam-1552	191	7	n	n	DET
ejpam-1552	191	8	miyamoto	miyamoto	NOUN
ejpam-1552	191	9	,	,	PUNCT
ejpam-1552	191	10	k	k	PROPN
ejpam-1552	191	11	yamamoto	yamamoto	PROPN
ejpam-1552	191	12	,	,	PUNCT
ejpam-1552	191	13	and	and	CCONJ
ejpam-1552	191	14	a	a	DET
ejpam-1552	191	15	sugiyama	sugiyama	NOUN
ejpam-1552	191	16	.	.	PUNCT
ejpam-1552	192	1	extensions	extension	NOUN
ejpam-1552	192	2	of	of	ADP
ejpam-1552	192	3	linear	linear	ADJ
ejpam-1552	192	4	diagonalparameter	diagonalparameter	NOUN
ejpam-1552	192	5	symmetry	symmetry	NOUN
ejpam-1552	192	6	and	and	CCONJ
ejpam-1552	192	7	quasi	quasi	ADJ
ejpam-1552	192	8	-	-	NOUN
ejpam-1552	192	9	symmetry	symmetry	ADJ
ejpam-1552	192	10	models	model	NOUN
ejpam-1552	192	11	for	for	ADP
ejpam-1552	192	12	cumulative	cumulative	ADJ
ejpam-1552	192	13	probabilities	probability	NOUN
ejpam-1552	192	14	in	in	ADP
ejpam-1552	192	15	square	square	ADJ
ejpam-1552	192	16	contingency	contingency	NOUN
ejpam-1552	192	17	tables	table	NOUN
ejpam-1552	192	18	.	.	PUNCT
ejpam-1552	193	1	statistica	statistica	PROPN
ejpam-1552	193	2	neerlandica	neerlandica	PROPN
ejpam-1552	193	3	,	,	PUNCT
ejpam-1552	193	4	61:273–283	61:273–283	PROPN
ejpam-1552	193	5	,	,	PUNCT
ejpam-1552	193	6	2007	2007	NUM
ejpam-1552	193	7	.	.	PUNCT
ejpam-1552	194	1	[	[	X
ejpam-1552	194	2	14	14	NUM
ejpam-1552	194	3	]	]	X
ejpam-1552	194	4	s	s	PART
ejpam-1552	194	5	tomizawa	tomizawa	PROPN
ejpam-1552	194	6	and	and	CCONJ
ejpam-1552	194	7	k	k	PROPN
ejpam-1552	194	8	tahata	tahata	PROPN
ejpam-1552	194	9	.	.	PUNCT
ejpam-1552	195	1	the	the	DET
ejpam-1552	195	2	analysis	analysis	NOUN
ejpam-1552	195	3	of	of	ADP
ejpam-1552	195	4	symmetry	symmetry	NOUN
ejpam-1552	195	5	and	and	CCONJ
ejpam-1552	195	6	asymmetry	asymmetry	NOUN
ejpam-1552	195	7	:	:	PUNCT
ejpam-1552	195	8	orthogonality	orthogonality	NOUN
ejpam-1552	195	9	of	of	ADP
ejpam-1552	195	10	decomposition	decomposition	NOUN
ejpam-1552	195	11	of	of	ADP
ejpam-1552	195	12	symmetry	symmetry	NOUN
ejpam-1552	195	13	into	into	ADP
ejpam-1552	195	14	quasi	quasi	NOUN
ejpam-1552	195	15	-	-	NOUN
ejpam-1552	195	16	symmetry	symmetry	ADJ
ejpam-1552	195	17	and	and	CCONJ
ejpam-1552	195	18	marginal	marginal	ADJ
ejpam-1552	195	19	symmetry	symmetry	NOUN
ejpam-1552	195	20	for	for	ADP
ejpam-1552	195	21	multi	multi	ADJ
ejpam-1552	195	22	-	-	ADJ
ejpam-1552	195	23	way	way	NOUN
ejpam-1552	195	24	tables	table	NOUN
ejpam-1552	195	25	.	.	PUNCT
ejpam-1552	196	1	journal	journal	PROPN
ejpam-1552	196	2	de	de	PROPN
ejpam-1552	196	3	la	la	PROPN
ejpam-1552	196	4	société	société	PROPN
ejpam-1552	196	5	française	française	PROPN
ejpam-1552	196	6	de	de	X
ejpam-1552	196	7	statistique	statistique	NOUN
ejpam-1552	196	8	,	,	PUNCT
ejpam-1552	196	9	148:3–36	148:3–36	NUM
ejpam-1552	196	10	,	,	PUNCT
ejpam-1552	196	11	2007	2007	NUM
ejpam-1552	196	12	.	.	PUNCT
ejpam-1552	197	1	[	[	X
ejpam-1552	197	2	15	15	NUM
ejpam-1552	197	3	]	]	X
ejpam-1552	197	4	h	h	NOUN
ejpam-1552	197	5	yamamoto	yamamoto	PROPN
ejpam-1552	197	6	,	,	PUNCT
ejpam-1552	197	7	t	t	PROPN
ejpam-1552	197	8	iwashita	iwashita	NOUN
ejpam-1552	197	9	,	,	PUNCT
ejpam-1552	197	10	and	and	CCONJ
ejpam-1552	197	11	s	s	VERB
ejpam-1552	197	12	tomizawa	tomizawa	NOUN
ejpam-1552	197	13	.	.	PUNCT
ejpam-1552	198	1	decomposition	decomposition	NOUN
ejpam-1552	198	2	of	of	ADP
ejpam-1552	198	3	symmetry	symmetry	NOUN
ejpam-1552	198	4	into	into	ADP
ejpam-1552	198	5	ordinal	ordinal	ADJ
ejpam-1552	198	6	quasi	quasi	NOUN
ejpam-1552	198	7	-	-	NOUN
ejpam-1552	198	8	symmetry	symmetry	ADJ
ejpam-1552	198	9	and	and	CCONJ
ejpam-1552	198	10	marginal	marginal	ADJ
ejpam-1552	198	11	equimoment	equimoment	NOUN
ejpam-1552	198	12	for	for	ADP
ejpam-1552	198	13	multi	multi	ADJ
ejpam-1552	198	14	-	-	ADJ
ejpam-1552	198	15	way	way	NOUN
ejpam-1552	198	16	tables	table	NOUN
ejpam-1552	198	17	.	.	PUNCT
ejpam-1552	199	1	austrian	austrian	ADJ
ejpam-1552	199	2	journal	journal	PROPN
ejpam-1552	199	3	of	of	ADP
ejpam-1552	199	4	statistics	statistic	NOUN
ejpam-1552	199	5	,	,	PUNCT
ejpam-1552	199	6	36:291–306	36:291–306	PROPN
ejpam-1552	199	7	,	,	PUNCT
ejpam-1552	199	8	2007	2007	NUM
ejpam-1552	199	9	.	.	PUNCT
ejpam-1552	200	1	[	[	X
ejpam-1552	200	2	16	16	NUM
ejpam-1552	200	3	]	]	X
ejpam-1552	200	4	k	k	PROPN
ejpam-1552	200	5	yamamoto	yamamoto	PROPN
ejpam-1552	200	6	,	,	PUNCT
ejpam-1552	200	7	s	s	PROPN
ejpam-1552	200	8	ando	ando	X
ejpam-1552	200	9	,	,	PUNCT
ejpam-1552	200	10	and	and	CCONJ
ejpam-1552	200	11	s	s	VERB
ejpam-1552	200	12	tomizawa	tomizawa	NOUN
ejpam-1552	200	13	.	.	PUNCT
ejpam-1552	201	1	decomposing	decompose	VERB
ejpam-1552	201	2	asymmetry	asymmetry	NOUN
ejpam-1552	201	3	into	into	ADP
ejpam-1552	201	4	extended	extended	ADJ
ejpam-1552	201	5	quasisymmetry	quasisymmetry	NOUN
ejpam-1552	201	6	and	and	CCONJ
ejpam-1552	201	7	marginal	marginal	ADJ
ejpam-1552	201	8	homogeneity	homogeneity	NOUN
ejpam-1552	201	9	for	for	ADP
ejpam-1552	201	10	cumulative	cumulative	ADJ
ejpam-1552	201	11	probabilities	probability	NOUN
ejpam-1552	201	12	in	in	ADP
ejpam-1552	201	13	square	square	ADJ
ejpam-1552	201	14	contingency	contingency	NOUN
ejpam-1552	201	15	tables	table	NOUN
ejpam-1552	201	16	.	.	PUNCT
ejpam-1552	202	1	journal	journal	PROPN
ejpam-1552	202	2	of	of	ADP
ejpam-1552	202	3	statistics	statistic	NOUN
ejpam-1552	202	4	:	:	PUNCT
ejpam-1552	202	5	advances	advance	NOUN
ejpam-1552	202	6	in	in	ADP
ejpam-1552	202	7	theory	theory	NOUN
ejpam-1552	202	8	and	and	CCONJ
ejpam-1552	202	9	applications	application	NOUN
ejpam-1552	202	10	,	,	PUNCT
ejpam-1552	202	11	5:1–13	5:1–13	NUM
ejpam-1552	202	12	,	,	PUNCT
ejpam-1552	202	13	2011	2011	NUM
ejpam-1552	202	14	.	.	PUNCT
ejpam-1552	203	1	[	[	X
ejpam-1552	203	2	17	17	NUM
ejpam-1552	203	3	]	]	X
ejpam-1552	203	4	k	k	PROPN
ejpam-1552	203	5	yamamoto	yamamoto	PROPN
ejpam-1552	203	6	and	and	CCONJ
ejpam-1552	203	7	s	s	PROPN
ejpam-1552	203	8	tomizawa	tomizawa	PROPN
ejpam-1552	203	9	.	.	PUNCT
ejpam-1552	204	1	analysis	analysis	NOUN
ejpam-1552	204	2	of	of	ADP
ejpam-1552	204	3	unaided	unaided	ADJ
ejpam-1552	204	4	vision	vision	NOUN
ejpam-1552	204	5	data	datum	NOUN
ejpam-1552	204	6	using	use	VERB
ejpam-1552	204	7	new	new	ADJ
ejpam-1552	204	8	decomposition	decomposition	NOUN
ejpam-1552	204	9	of	of	ADP
ejpam-1552	204	10	symmetry	symmetry	NOUN
ejpam-1552	204	11	.	.	PUNCT
ejpam-1552	205	1	american	american	PROPN
ejpam-1552	205	2	medical	medical	PROPN
ejpam-1552	205	3	journal	journal	PROPN
ejpam-1552	205	4	,	,	PUNCT
ejpam-1552	205	5	3:37–42	3:37–42	NUM
ejpam-1552	205	6	,	,	PUNCT
ejpam-1552	205	7	2012	2012	NUM
ejpam-1552	205	8	.	.	PUNCT