id	sid	tid	token	lemma	pos
ajis02-348	1	1	30	30	NUM
ajis02-348	1	2	assessing	assess	VERB
ajis02-348	1	3	software	software	NOUN
ajis02-348	1	4	cost	cost	NOUN
ajis02-348	1	5	estimation	estimation	NOUN
ajis02-348	1	6	models	model	NOUN
ajis02-348	1	7	:	:	PUNCT
ajis02-348	1	8	criteria	criterion	NOUN
ajis02-348	1	9	for	for	ADP
ajis02-348	1	10	accuracy	accuracy	NOUN
ajis02-348	1	11	,	,	PUNCT
ajis02-348	1	12	consistency	consistency	NOUN
ajis02-348	1	13	and	and	CCONJ
ajis02-348	1	14	regression	regression	VERB
ajis02-348	1	15	bruce	bruce	PROPN
ajis02-348	1	16	w.n	w.n	PROPN
ajis02-348	1	17	.	.	PROPN
ajis02-348	1	18	lo	lo	PROPN
ajis02-348	1	19	and	and	CCONJ
ajis02-348	1	20	xiangzhu	xiangzhu	PROPN
ajis02-348	1	21	gao	gao	PROPN
ajis02-348	1	22	school	school	PROPN
ajis02-348	1	23	of	of	ADP
ajis02-348	1	24	multimedia	multimedia	NOUN
ajis02-348	1	25	and	and	CCONJ
ajis02-348	1	26	information	information	NOUN
ajis02-348	1	27	technology	technology	PROPN
ajis02-348	1	28	southern	southern	PROPN
ajis02-348	1	29	cross	cross	PROPN
ajis02-348	1	30	university	university	PROPN
ajis02-348	1	31	lismore	lismore	PROPN
ajis02-348	1	32	,	,	PUNCT
ajis02-348	1	33	nsw	nsw	PROPN
ajis02-348	1	34	australia	australia	PROPN
ajis02-348	1	35	2480	2480	NUM
ajis02-348	1	36	email	email	NOUN
ajis02-348	1	37	:	:	PUNCT
ajis02-348	1	38	blo@scu.edu.au	blo@scu.edu.au	PROPN
ajis02-348	1	39	abstract	abstract	VERB
ajis02-348	1	40	one	one	NUM
ajis02-348	1	41	of	of	ADP
ajis02-348	1	42	the	the	DET
ajis02-348	1	43	problems	problem	NOUN
ajis02-348	1	44	in	in	ADP
ajis02-348	1	45	software	software	NOUN
ajis02-348	1	46	cost	cost	NOUN
ajis02-348	1	47	estimation	estimation	NOUN
ajis02-348	1	48	is	be	AUX
ajis02-348	1	49	how	how	SCONJ
ajis02-348	1	50	to	to	PART
ajis02-348	1	51	evaluate	evaluate	VERB
ajis02-348	1	52	estimation	estimation	NOUN
ajis02-348	1	53	models	model	NOUN
ajis02-348	1	54	.	.	PUNCT
ajis02-348	2	1	estimation	estimation	NOUN
ajis02-348	2	2	models	model	NOUN
ajis02-348	2	3	are	be	AUX
ajis02-348	2	4	usually	usually	ADV
ajis02-348	2	5	evaluated	evaluate	VERB
ajis02-348	2	6	against	against	ADP
ajis02-348	2	7	two	two	NUM
ajis02-348	2	8	attributes	attribute	NOUN
ajis02-348	2	9	:	:	PUNCT
ajis02-348	2	10	estimation	estimation	NOUN
ajis02-348	2	11	accuracy	accuracy	NOUN
ajis02-348	2	12	and	and	CCONJ
ajis02-348	2	13	estimation	estimation	NOUN
ajis02-348	2	14	consistency	consistency	NOUN
ajis02-348	2	15	.	.	PUNCT
ajis02-348	3	1	a	a	DET
ajis02-348	3	2	number	number	NOUN
ajis02-348	3	3	of	of	ADP
ajis02-348	3	4	measures	measure	NOUN
ajis02-348	3	5	are	be	AUX
ajis02-348	3	6	reported	report	VERB
ajis02-348	3	7	in	in	ADP
ajis02-348	3	8	the	the	DET
ajis02-348	3	9	literature	literature	NOUN
ajis02-348	3	10	,	,	PUNCT
ajis02-348	3	11	but	but	CCONJ
ajis02-348	3	12	they	they	PRON
ajis02-348	3	13	have	have	VERB
ajis02-348	3	14	shortcomings	shortcoming	NOUN
ajis02-348	3	15	.	.	PUNCT
ajis02-348	4	1	there	there	PRON
ajis02-348	4	2	is	be	VERB
ajis02-348	4	3	no	no	DET
ajis02-348	4	4	generally	generally	ADV
ajis02-348	4	5	accepted	accept	VERB
ajis02-348	4	6	standard	standard	NOUN
ajis02-348	4	7	to	to	PART
ajis02-348	4	8	evaluate	evaluate	VERB
ajis02-348	4	9	estimation	estimation	NOUN
ajis02-348	4	10	models	model	NOUN
ajis02-348	4	11	and	and	CCONJ
ajis02-348	4	12	the	the	DET
ajis02-348	4	13	existing	exist	VERB
ajis02-348	4	14	measures	measure	NOUN
ajis02-348	4	15	sometimes	sometimes	ADV
ajis02-348	4	16	are	be	AUX
ajis02-348	4	17	not	not	PART
ajis02-348	4	18	consistent	consistent	ADJ
ajis02-348	4	19	among	among	ADP
ajis02-348	4	20	themselves	themselves	PRON
ajis02-348	4	21	.	.	PUNCT
ajis02-348	5	1	this	this	DET
ajis02-348	5	2	paper	paper	NOUN
ajis02-348	5	3	examines	examine	VERB
ajis02-348	5	4	existing	exist	VERB
ajis02-348	5	5	measures	measure	NOUN
ajis02-348	5	6	of	of	ADP
ajis02-348	5	7	estimation	estimation	NOUN
ajis02-348	5	8	accuracy	accuracy	NOUN
ajis02-348	5	9	and	and	CCONJ
ajis02-348	5	10	consistency	consistency	NOUN
ajis02-348	5	11	and	and	CCONJ
ajis02-348	5	12	proposes	propose	VERB
ajis02-348	5	13	two	two	NUM
ajis02-348	5	14	new	new	ADJ
ajis02-348	5	15	ones	one	NOUN
ajis02-348	5	16	:	:	PUNCT
ajis02-348	5	17	the	the	DET
ajis02-348	5	18	weighted	weight	VERB
ajis02-348	5	19	mean	mean	NOUN
ajis02-348	5	20	of	of	ADP
ajis02-348	5	21	quartiles	quartile	NOUN
ajis02-348	5	22	of	of	ADP
ajis02-348	5	23	relative	relative	ADJ
ajis02-348	5	24	errors	error	NOUN
ajis02-348	5	25	(	(	PUNCT
ajis02-348	5	26	wmq	wmq	PROPN
ajis02-348	5	27	)	)	PUNCT
ajis02-348	5	28	as	as	ADP
ajis02-348	5	29	a	a	DET
ajis02-348	5	30	measure	measure	NOUN
ajis02-348	5	31	of	of	ADP
ajis02-348	5	32	accuracy	accuracy	NOUN
ajis02-348	5	33	and	and	CCONJ
ajis02-348	5	34	the	the	DET
ajis02-348	5	35	standard	standard	ADJ
ajis02-348	5	36	deviation	deviation	NOUN
ajis02-348	5	37	of	of	ADP
ajis02-348	5	38	the	the	DET
ajis02-348	5	39	ratios	ratio	NOUN
ajis02-348	5	40	of	of	ADP
ajis02-348	5	41	the	the	DET
ajis02-348	5	42	estimate	estimate	NOUN
ajis02-348	5	43	to	to	ADP
ajis02-348	5	44	actual	actual	ADJ
ajis02-348	5	45	observation	observation	NOUN
ajis02-348	5	46	(	(	PUNCT
ajis02-348	5	47	sdr	sdr	NOUN
ajis02-348	5	48	)	)	PUNCT
ajis02-348	5	49	as	as	ADP
ajis02-348	5	50	a	a	DET
ajis02-348	5	51	measure	measure	NOUN
ajis02-348	5	52	of	of	ADP
ajis02-348	5	53	consistency	consistency	NOUN
ajis02-348	5	54	.	.	PUNCT
ajis02-348	6	1	besides	besides	SCONJ
ajis02-348	6	2	,	,	PUNCT
ajis02-348	6	3	a	a	DET
ajis02-348	6	4	new	new	ADJ
ajis02-348	6	5	regression	regression	NOUN
ajis02-348	6	6	criterion	criterion	NOUN
ajis02-348	6	7	is	be	AUX
ajis02-348	6	8	proposed	propose	VERB
ajis02-348	6	9	to	to	PART
ajis02-348	6	10	determine	determine	VERB
ajis02-348	6	11	model	model	NOUN
ajis02-348	6	12	parameters	parameter	NOUN
ajis02-348	6	13	.	.	PUNCT
ajis02-348	7	1	the	the	DET
ajis02-348	7	2	proposed	propose	VERB
ajis02-348	7	3	measures	measure	NOUN
ajis02-348	7	4	and	and	CCONJ
ajis02-348	7	5	criterion	criterion	NOUN
ajis02-348	7	6	were	be	AUX
ajis02-348	7	7	tested	test	VERB
ajis02-348	7	8	with	with	ADP
ajis02-348	7	9	a	a	DET
ajis02-348	7	10	data	datum	NOUN
ajis02-348	7	11	set	set	VERB
ajis02-348	7	12	from	from	ADP
ajis02-348	7	13	real	real	ADJ
ajis02-348	7	14	world	world	NOUN
ajis02-348	7	15	software	software	NOUN
ajis02-348	7	16	projects	project	NOUN
ajis02-348	7	17	.	.	PUNCT
ajis02-348	8	1	results	result	NOUN
ajis02-348	8	2	obtained	obtain	VERB
ajis02-348	8	3	show	show	VERB
ajis02-348	8	4	that	that	SCONJ
ajis02-348	8	5	these	these	DET
ajis02-348	8	6	new	new	ADJ
ajis02-348	8	7	measures	measure	NOUN
ajis02-348	8	8	and	and	CCONJ
ajis02-348	8	9	criterion	criterion	NOUN
ajis02-348	8	10	overcome	overcome	VERB
ajis02-348	8	11	many	many	ADJ
ajis02-348	8	12	of	of	ADP
ajis02-348	8	13	the	the	DET
ajis02-348	8	14	difficulties	difficulty	NOUN
ajis02-348	8	15	of	of	ADP
ajis02-348	8	16	the	the	DET
ajis02-348	8	17	existing	exist	VERB
ajis02-348	8	18	ones	one	NOUN
ajis02-348	8	19	.	.	PUNCT
ajis02-348	9	1	introduction	introduction	NOUN
ajis02-348	9	2	since	since	SCONJ
ajis02-348	9	3	the	the	DET
ajis02-348	9	4	early	early	ADJ
ajis02-348	9	5	1950s	1950s	NUM
ajis02-348	9	6	,	,	PUNCT
ajis02-348	9	7	software	software	NOUN
ajis02-348	9	8	development	development	NOUN
ajis02-348	9	9	practitioners	practitioner	NOUN
ajis02-348	9	10	and	and	CCONJ
ajis02-348	9	11	researchers	researcher	NOUN
ajis02-348	9	12	have	have	AUX
ajis02-348	9	13	been	be	AUX
ajis02-348	9	14	trying	try	VERB
ajis02-348	9	15	to	to	PART
ajis02-348	9	16	develop	develop	VERB
ajis02-348	9	17	methods	method	NOUN
ajis02-348	9	18	to	to	PART
ajis02-348	9	19	estimate	estimate	VERB
ajis02-348	9	20	software	software	NOUN
ajis02-348	9	21	costs	cost	NOUN
ajis02-348	9	22	and	and	CCONJ
ajis02-348	9	23	schedules	schedule	NOUN
ajis02-348	9	24	(	(	PUNCT
ajis02-348	9	25	abdel	abdel	PROPN
ajis02-348	9	26	-	-	PUNCT
ajis02-348	9	27	hamid	hamid	PROPN
ajis02-348	9	28	,	,	PUNCT
ajis02-348	9	29	1990	1990	NUM
ajis02-348	9	30	)	)	PUNCT
ajis02-348	9	31	.	.	PUNCT
ajis02-348	10	1	software	software	NOUN
ajis02-348	10	2	cost	cost	NOUN
ajis02-348	10	3	estimation	estimation	NOUN
ajis02-348	10	4	models	model	NOUN
ajis02-348	10	5	have	have	AUX
ajis02-348	10	6	appeared	appear	VERB
ajis02-348	10	7	in	in	ADP
ajis02-348	10	8	the	the	DET
ajis02-348	10	9	literature	literature	NOUN
ajis02-348	10	10	over	over	ADP
ajis02-348	10	11	the	the	DET
ajis02-348	10	12	past	past	ADJ
ajis02-348	10	13	two	two	NUM
ajis02-348	10	14	decades	decade	NOUN
ajis02-348	10	15	(	(	PUNCT
ajis02-348	10	16	wrigley	wrigley	PROPN
ajis02-348	10	17	et	et	PROPN
ajis02-348	10	18	al	al	PROPN
ajis02-348	10	19	,	,	PUNCT
ajis02-348	10	20	1991	1991	NUM
ajis02-348	10	21	)	)	PUNCT
ajis02-348	10	22	.	.	PUNCT
ajis02-348	11	1	however	however	ADV
ajis02-348	11	2	,	,	PUNCT
ajis02-348	11	3	the	the	DET
ajis02-348	11	4	field	field	NOUN
ajis02-348	11	5	of	of	ADP
ajis02-348	11	6	software	software	NOUN
ajis02-348	11	7	cost	cost	NOUN
ajis02-348	11	8	estimation	estimation	NOUN
ajis02-348	11	9	is	be	AUX
ajis02-348	11	10	still	still	ADV
ajis02-348	11	11	in	in	ADP
ajis02-348	11	12	its	its	PRON
ajis02-348	11	13	infancy	infancy	NOUN
ajis02-348	11	14	(	(	PUNCT
ajis02-348	11	15	kitchenham	kitchenham	INTJ
ajis02-348	11	16	et	et	PROPN
ajis02-348	11	17	al	al	PROPN
ajis02-348	11	18	,	,	PUNCT
ajis02-348	11	19	1990	1990	NUM
ajis02-348	11	20	)	)	PUNCT
ajis02-348	11	21	.	.	PUNCT
ajis02-348	12	1	the	the	DET
ajis02-348	12	2	existing	exist	VERB
ajis02-348	12	3	cost	cost	NOUN
ajis02-348	12	4	estimation	estimation	NOUN
ajis02-348	12	5	methods	method	NOUN
ajis02-348	12	6	are	be	AUX
ajis02-348	12	7	far	far	ADV
ajis02-348	12	8	from	from	ADP
ajis02-348	12	9	standardised	standardised	ADJ
ajis02-348	12	10	and	and	CCONJ
ajis02-348	12	11	reliable	reliable	ADJ
ajis02-348	12	12	(	(	PUNCT
ajis02-348	12	13	rowlands	rowland	NOUN
ajis02-348	12	14	,	,	PUNCT
ajis02-348	12	15	1989	1989	NUM
ajis02-348	12	16	)	)	PUNCT
ajis02-348	12	17	.	.	PUNCT
ajis02-348	13	1	there	there	PRON
ajis02-348	13	2	is	be	VERB
ajis02-348	13	3	a	a	DET
ajis02-348	13	4	need	need	NOUN
ajis02-348	13	5	to	to	PART
ajis02-348	13	6	evaluate	evaluate	VERB
ajis02-348	13	7	estimation	estimation	NOUN
ajis02-348	13	8	models	model	NOUN
ajis02-348	13	9	and	and	CCONJ
ajis02-348	13	10	improve	improve	VERB
ajis02-348	13	11	modeling	modeling	NOUN
ajis02-348	13	12	processes	process	NOUN
ajis02-348	13	13	.	.	PUNCT
ajis02-348	14	1	this	this	DET
ajis02-348	14	2	paper	paper	NOUN
ajis02-348	14	3	will	will	AUX
ajis02-348	14	4	focus	focus	VERB
ajis02-348	14	5	on	on	ADP
ajis02-348	14	6	how	how	SCONJ
ajis02-348	14	7	to	to	PART
ajis02-348	14	8	quantitatively	quantitatively	ADV
ajis02-348	14	9	evaluate	evaluate	VERB
ajis02-348	14	10	software	software	NOUN
ajis02-348	14	11	cost	cost	NOUN
ajis02-348	14	12	estimation	estimation	NOUN
ajis02-348	14	13	models	model	NOUN
ajis02-348	14	14	.	.	PUNCT
ajis02-348	15	1	a	a	DET
ajis02-348	15	2	new	new	ADJ
ajis02-348	15	3	approach	approach	NOUN
ajis02-348	15	4	to	to	PART
ajis02-348	15	5	determine	determine	VERB
ajis02-348	15	6	model	model	NOUN
ajis02-348	15	7	parameters	parameter	NOUN
ajis02-348	15	8	is	be	AUX
ajis02-348	15	9	also	also	ADV
ajis02-348	15	10	proposed	propose	VERB
ajis02-348	15	11	.	.	PUNCT
ajis02-348	16	1	in	in	ADP
ajis02-348	16	2	the	the	DET
ajis02-348	16	3	field	field	NOUN
ajis02-348	16	4	of	of	ADP
ajis02-348	16	5	software	software	NOUN
ajis02-348	16	6	cost	cost	NOUN
ajis02-348	16	7	estimation	estimation	NOUN
ajis02-348	16	8	,	,	PUNCT
ajis02-348	16	9	estimation	estimation	NOUN
ajis02-348	16	10	models	model	NOUN
ajis02-348	16	11	are	be	AUX
ajis02-348	16	12	usually	usually	ADV
ajis02-348	16	13	evaluated	evaluate	VERB
ajis02-348	16	14	against	against	ADP
ajis02-348	16	15	two	two	NUM
ajis02-348	16	16	attributes	attribute	NOUN
ajis02-348	16	17	:	:	PUNCT
ajis02-348	16	18	estimation	estimation	NOUN
ajis02-348	16	19	accuracy	accuracy	NOUN
ajis02-348	16	20	and	and	CCONJ
ajis02-348	16	21	estimation	estimation	NOUN
ajis02-348	16	22	consistency	consistency	NOUN
ajis02-348	16	23	.	.	PUNCT
ajis02-348	17	1	the	the	DET
ajis02-348	17	2	rules	rule	NOUN
ajis02-348	17	3	or	or	CCONJ
ajis02-348	17	4	measures	measure	NOUN
ajis02-348	17	5	needed	need	VERB
ajis02-348	17	6	to	to	PART
ajis02-348	17	7	describe	describe	VERB
ajis02-348	17	8	these	these	DET
ajis02-348	17	9	two	two	NUM
ajis02-348	17	10	attributes	attribute	NOUN
ajis02-348	17	11	will	will	AUX
ajis02-348	17	12	be	be	AUX
ajis02-348	17	13	discussed	discuss	VERB
ajis02-348	17	14	in	in	ADP
ajis02-348	17	15	this	this	DET
ajis02-348	17	16	paper	paper	NOUN
ajis02-348	17	17	.	.	PUNCT
ajis02-348	18	1	measurement	measurement	NOUN
ajis02-348	18	2	may	may	AUX
ajis02-348	18	3	be	be	AUX
ajis02-348	18	4	used	use	VERB
ajis02-348	18	5	for	for	ADP
ajis02-348	18	6	two	two	NUM
ajis02-348	18	7	purposes	purpose	NOUN
ajis02-348	18	8	:	:	PUNCT
ajis02-348	18	9	assessment	assessment	NOUN
ajis02-348	18	10	and	and	CCONJ
ajis02-348	18	11	prediction	prediction	NOUN
ajis02-348	18	12	(	(	PUNCT
ajis02-348	18	13	fenton	fenton	NOUN
ajis02-348	18	14	,	,	PUNCT
ajis02-348	18	15	1994	1994	NUM
ajis02-348	18	16	)	)	PUNCT
ajis02-348	18	17	.	.	PUNCT
ajis02-348	19	1	for	for	ADP
ajis02-348	19	2	assessment	assessment	NOUN
ajis02-348	19	3	,	,	PUNCT
ajis02-348	19	4	the	the	DET
ajis02-348	19	5	results	result	NOUN
ajis02-348	19	6	of	of	ADP
ajis02-348	19	7	comparing	compare	VERB
ajis02-348	19	8	different	different	ADJ
ajis02-348	19	9	models	model	NOUN
ajis02-348	19	10	may	may	AUX
ajis02-348	19	11	be	be	AUX
ajis02-348	19	12	used	use	VERB
ajis02-348	19	13	to	to	PART
ajis02-348	19	14	judge	judge	VERB
ajis02-348	19	15	which	which	DET
ajis02-348	19	16	theory	theory	NOUN
ajis02-348	19	17	is	be	AUX
ajis02-348	19	18	more	more	ADV
ajis02-348	19	19	successful	successful	ADJ
ajis02-348	19	20	at	at	ADP
ajis02-348	19	21	explaining	explain	VERB
ajis02-348	19	22	the	the	DET
ajis02-348	19	23	behaviour	behaviour	NOUN
ajis02-348	19	24	of	of	ADP
ajis02-348	19	25	cost	cost	NOUN
ajis02-348	19	26	factors	factor	NOUN
ajis02-348	19	27	(	(	PUNCT
ajis02-348	19	28	pfleeger	pfleeger	NOUN
ajis02-348	19	29	,	,	PUNCT
ajis02-348	19	30	1991	1991	NUM
ajis02-348	19	31	)	)	PUNCT
ajis02-348	19	32	.	.	PUNCT
ajis02-348	20	1	successful	successful	ADJ
ajis02-348	20	2	models	model	NOUN
ajis02-348	20	3	would	would	AUX
ajis02-348	20	4	provide	provide	VERB
ajis02-348	20	5	further	further	ADJ
ajis02-348	20	6	insight	insight	NOUN
ajis02-348	20	7	into	into	ADP
ajis02-348	20	8	software	software	NOUN
ajis02-348	20	9	development	development	NOUN
ajis02-348	20	10	processes	process	NOUN
ajis02-348	20	11	.	.	PUNCT
ajis02-348	21	1	to	to	PART
ajis02-348	21	2	assess	assess	VERB
ajis02-348	21	3	models	model	NOUN
ajis02-348	21	4	,	,	PUNCT
ajis02-348	21	5	we	we	PRON
ajis02-348	21	6	need	need	VERB
ajis02-348	21	7	to	to	PART
ajis02-348	21	8	have	have	VERB
ajis02-348	21	9	common	common	ADJ
ajis02-348	21	10	standards	standard	NOUN
ajis02-348	21	11	.	.	PUNCT
ajis02-348	22	1	although	although	SCONJ
ajis02-348	22	2	some	some	DET
ajis02-348	22	3	measures	measure	NOUN
ajis02-348	22	4	for	for	ADP
ajis02-348	22	5	estimation	estimation	NOUN
ajis02-348	22	6	accuracy	accuracy	NOUN
ajis02-348	22	7	and	and	CCONJ
ajis02-348	22	8	consistency	consistency	NOUN
ajis02-348	22	9	have	have	AUX
ajis02-348	22	10	been	be	AUX
ajis02-348	22	11	introduced	introduce	VERB
ajis02-348	22	12	in	in	ADP
ajis02-348	22	13	the	the	DET
ajis02-348	22	14	literature	literature	NOUN
ajis02-348	22	15	,	,	PUNCT
ajis02-348	22	16	they	they	PRON
ajis02-348	22	17	are	be	AUX
ajis02-348	22	18	not	not	PART
ajis02-348	22	19	generally	generally	ADV
ajis02-348	22	20	accepted	accept	VERB
ajis02-348	22	21	for	for	ADP
ajis02-348	22	22	model	model	NOUN
ajis02-348	22	23	assessment	assessment	NOUN
ajis02-348	22	24	(	(	PUNCT
ajis02-348	22	25	verner	verner	NOUN
ajis02-348	22	26	et	et	PROPN
ajis02-348	22	27	al	al	PROPN
ajis02-348	22	28	,	,	PUNCT
ajis02-348	22	29	1992	1992	NUM
ajis02-348	22	30	)	)	PUNCT
ajis02-348	22	31	.	.	PUNCT
ajis02-348	23	1	sometimes	sometimes	ADV
ajis02-348	23	2	these	these	DET
ajis02-348	23	3	measures	measure	NOUN
ajis02-348	23	4	are	be	AUX
ajis02-348	23	5	not	not	PART
ajis02-348	23	6	consistent	consistent	ADJ
ajis02-348	23	7	among	among	ADP
ajis02-348	23	8	themselves	themselves	PRON
ajis02-348	23	9	.	.	PUNCT
ajis02-348	24	1	for	for	ADP
ajis02-348	24	2	example	example	NOUN
ajis02-348	24	3	,	,	PUNCT
ajis02-348	24	4	a	a	DET
ajis02-348	24	5	model	model	NOUN
ajis02-348	24	6	may	may	AUX
ajis02-348	24	7	be	be	AUX
ajis02-348	24	8	better	well	ADJ
ajis02-348	24	9	than	than	ADP
ajis02-348	24	10	another	another	DET
ajis02-348	24	11	one	one	NOUN
ajis02-348	24	12	with	with	ADP
ajis02-348	24	13	respect	respect	NOUN
ajis02-348	24	14	to	to	ADP
ajis02-348	24	15	one	one	NUM
ajis02-348	24	16	measure	measure	NOUN
ajis02-348	24	17	but	but	CCONJ
ajis02-348	24	18	poorer	poor	ADJ
ajis02-348	24	19	with	with	ADP
ajis02-348	24	20	respect	respect	NOUN
ajis02-348	24	21	to	to	ADP
ajis02-348	24	22	a	a	DET
ajis02-348	24	23	different	different	ADJ
ajis02-348	24	24	measure	measure	NOUN
ajis02-348	24	25	.	.	PUNCT
ajis02-348	25	1	for	for	ADP
ajis02-348	25	2	prediction	prediction	NOUN
ajis02-348	25	3	,	,	PUNCT
ajis02-348	25	4	the	the	DET
ajis02-348	25	5	measurement	measurement	NOUN
ajis02-348	25	6	may	may	AUX
ajis02-348	25	7	provide	provide	VERB
ajis02-348	25	8	feedback	feedback	NOUN
ajis02-348	25	9	to	to	PART
ajis02-348	25	10	improve	improve	VERB
ajis02-348	25	11	the	the	DET
ajis02-348	25	12	modeling	modeling	NOUN
ajis02-348	25	13	process	process	NOUN
ajis02-348	25	14	so	so	SCONJ
ajis02-348	25	15	that	that	SCONJ
ajis02-348	25	16	the	the	DET
ajis02-348	25	17	model	model	NOUN
ajis02-348	25	18	can	can	AUX
ajis02-348	25	19	satisfy	satisfy	VERB
ajis02-348	25	20	the	the	DET
ajis02-348	25	21	rules	rule	NOUN
ajis02-348	25	22	of	of	ADP
ajis02-348	25	23	measurement	measurement	NOUN
ajis02-348	25	24	as	as	ADV
ajis02-348	25	25	far	far	ADV
ajis02-348	25	26	as	as	ADP
ajis02-348	25	27	possible	possible	ADJ
ajis02-348	25	28	.	.	PUNCT
ajis02-348	26	1	modeling	modeling	NOUN
ajis02-348	26	2	is	be	AUX
ajis02-348	26	3	associated	associate	VERB
ajis02-348	26	4	with	with	ADP
ajis02-348	26	5	measurement	measurement	NOUN
ajis02-348	26	6	(	(	PUNCT
ajis02-348	26	7	kan	kan	PROPN
ajis02-348	26	8	et	et	PROPN
ajis02-348	26	9	al	al	PROPN
ajis02-348	26	10	,	,	PUNCT
ajis02-348	26	11	1994	1994	NUM
ajis02-348	26	12	)	)	PUNCT
ajis02-348	26	13	.	.	PUNCT
ajis02-348	27	1	in	in	ADP
ajis02-348	27	2	this	this	DET
ajis02-348	27	3	context	context	NOUN
ajis02-348	27	4	,	,	PUNCT
ajis02-348	27	5	it	it	PRON
ajis02-348	27	6	is	be	AUX
ajis02-348	27	7	necessary	necessary	ADJ
ajis02-348	27	8	to	to	PART
ajis02-348	27	9	define	define	VERB
ajis02-348	27	10	the	the	DET
ajis02-348	27	11	procedures	procedure	NOUN
ajis02-348	27	12	for	for	ADP
ajis02-348	27	13	determining	determine	VERB
ajis02-348	27	14	model	model	NOUN
ajis02-348	27	15	parameters	parameter	NOUN
ajis02-348	27	16	and	and	CCONJ
ajis02-348	27	17	interpreting	interpret	VERB
ajis02-348	27	18	the	the	DET
ajis02-348	27	19	results	result	NOUN
ajis02-348	27	20	(	(	PUNCT
ajis02-348	27	21	fenton	fenton	NOUN
ajis02-348	27	22	,	,	PUNCT
ajis02-348	27	23	1994	1994	NUM
ajis02-348	27	24	)	)	PUNCT
ajis02-348	27	25	.	.	PUNCT
ajis02-348	28	1	this	this	PRON
ajis02-348	28	2	leads	lead	VERB
ajis02-348	28	3	to	to	ADP
ajis02-348	28	4	the	the	DET
ajis02-348	28	5	problem	problem	NOUN
ajis02-348	28	6	of	of	ADP
ajis02-348	28	7	how	how	SCONJ
ajis02-348	28	8	to	to	PART
ajis02-348	28	9	determine	determine	VERB
ajis02-348	28	10	the	the	DET
ajis02-348	28	11	model	model	NOUN
ajis02-348	28	12	parameters	parameter	NOUN
ajis02-348	28	13	.	.	PUNCT
ajis02-348	29	1	unfortunately	unfortunately	ADV
ajis02-348	29	2	there	there	PRON
ajis02-348	29	3	is	be	VERB
ajis02-348	29	4	also	also	ADV
ajis02-348	29	5	no	no	DET
ajis02-348	29	6	generally	generally	ADV
ajis02-348	29	7	accepted	accept	VERB
ajis02-348	29	8	criterion	criterion	NOUN
ajis02-348	29	9	for	for	SCONJ
ajis02-348	29	10	researchers	researcher	NOUN
ajis02-348	29	11	to	to	PART
ajis02-348	29	12	follow	follow	VERB
ajis02-348	29	13	in	in	ADP
ajis02-348	29	14	the	the	DET
ajis02-348	29	15	modelling	modelling	NOUN
ajis02-348	29	16	process	process	NOUN
ajis02-348	29	17	.	.	PUNCT
ajis02-348	30	1	this	this	DET
ajis02-348	30	2	paper	paper	NOUN
ajis02-348	30	3	,	,	PUNCT
ajis02-348	30	4	therefore	therefore	ADV
ajis02-348	30	5	,	,	PUNCT
ajis02-348	30	6	will	will	AUX
ajis02-348	30	7	also	also	ADV
ajis02-348	30	8	discuss	discuss	VERB
ajis02-348	30	9	the	the	DET
ajis02-348	30	10	criteria	criterion	NOUN
ajis02-348	30	11	for	for	ADP
ajis02-348	30	12	determining	determine	VERB
ajis02-348	30	13	model	model	NOUN
ajis02-348	30	14	parameters	parameter	NOUN
ajis02-348	30	15	.	.	PUNCT
ajis02-348	31	1	the	the	DET
ajis02-348	31	2	paper	paper	NOUN
ajis02-348	31	3	is	be	AUX
ajis02-348	31	4	organised	organise	VERB
ajis02-348	31	5	into	into	ADP
ajis02-348	31	6	three	three	NUM
ajis02-348	31	7	parts	part	NOUN
ajis02-348	31	8	.	.	PUNCT
ajis02-348	32	1	the	the	DET
ajis02-348	32	2	first	first	ADJ
ajis02-348	32	3	part	part	NOUN
ajis02-348	32	4	examines	examine	VERB
ajis02-348	32	5	the	the	DET
ajis02-348	32	6	measures	measure	NOUN
ajis02-348	32	7	used	use	VERB
ajis02-348	32	8	to	to	PART
ajis02-348	32	9	evaluate	evaluate	VERB
ajis02-348	32	10	estimation	estimation	NOUN
ajis02-348	32	11	models	model	NOUN
ajis02-348	32	12	.	.	PUNCT
ajis02-348	33	1	to	to	PART
ajis02-348	33	2	overcome	overcome	VERB
ajis02-348	33	3	the	the	DET
ajis02-348	33	4	shortcomings	shortcoming	NOUN
ajis02-348	33	5	of	of	ADP
ajis02-348	33	6	existing	exist	VERB
ajis02-348	33	7	practice	practice	NOUN
ajis02-348	33	8	,	,	PUNCT
ajis02-348	33	9	two	two	NUM
ajis02-348	33	10	new	new	ADJ
ajis02-348	33	11	measures	measure	NOUN
ajis02-348	33	12	are	be	AUX
ajis02-348	33	13	proposed	propose	VERB
ajis02-348	33	14	:	:	PUNCT
ajis02-348	33	15	the	the	DET
ajis02-348	33	16	weighted	weight	VERB
ajis02-348	33	17	mean	mean	NOUN
ajis02-348	33	18	of	of	ADP
ajis02-348	33	19	quartiles	quartile	NOUN
ajis02-348	33	20	of	of	ADP
ajis02-348	33	21	relative	relative	ADJ
ajis02-348	33	22	errors	error	NOUN
ajis02-348	33	23	(	(	PUNCT
ajis02-348	33	24	wmq	wmq	PROPN
ajis02-348	33	25	)	)	PUNCT
ajis02-348	33	26	which	which	PRON
ajis02-348	33	27	provides	provide	VERB
ajis02-348	33	28	a	a	DET
ajis02-348	33	29	measurement	measurement	NOUN
ajis02-348	33	30	of	of	ADP
ajis02-348	33	31	accuracy	accuracy	NOUN
ajis02-348	33	32	and	and	CCONJ
ajis02-348	33	33	the	the	DET
ajis02-348	33	34	standard	standard	ADJ
ajis02-348	33	35	deviation	deviation	NOUN
ajis02-348	33	36	of	of	ADP
ajis02-348	33	37	the	the	DET
ajis02-348	33	38	ratios	ratio	NOUN
ajis02-348	33	39	of	of	ADP
ajis02-348	33	40	the	the	DET
ajis02-348	33	41	estimate	estimate	NOUN
ajis02-348	33	42	to	to	ADP
ajis02-348	33	43	actual	actual	ADJ
ajis02-348	33	44	value	value	NOUN
ajis02-348	33	45	(	(	PUNCT
ajis02-348	33	46	sdr	sdr	PROPN
ajis02-348	33	47	)	)	PUNCT
ajis02-348	33	48	which	which	PRON
ajis02-348	33	49	provides	provide	VERB
ajis02-348	33	50	a	a	DET
ajis02-348	33	51	measurement	measurement	NOUN
ajis02-348	33	52	of	of	ADP
ajis02-348	33	53	consistency	consistency	NOUN
ajis02-348	33	54	.	.	PUNCT
ajis02-348	34	1	the	the	DET
ajis02-348	34	2	second	second	ADJ
ajis02-348	34	3	part	part	NOUN
ajis02-348	34	4	examines	examine	VERB
ajis02-348	34	5	traditional	traditional	ADJ
ajis02-348	34	6	mathematical	mathematical	ADJ
ajis02-348	34	7	procedures	procedure	NOUN
ajis02-348	34	8	for	for	ADP
ajis02-348	34	9	formulating	formulate	VERB
ajis02-348	34	10	costing	costing	NOUN
ajis02-348	34	11	models	model	NOUN
ajis02-348	34	12	and	and	CCONJ
ajis02-348	34	13	proposes	propose	VERB
ajis02-348	34	14	a	a	DET
ajis02-348	34	15	new	new	ADJ
ajis02-348	34	16	regression	regression	NOUN
ajis02-348	34	17	criterion	criterion	NOUN
ajis02-348	34	18	called	call	VERB
ajis02-348	34	19	least	least	ADJ
ajis02-348	34	20	sum	sum	NOUN
ajis02-348	34	21	of	of	ADP
ajis02-348	34	22	logarithmic	logarithmic	ADJ
ajis02-348	34	23	ratios	ratio	NOUN
ajis02-348	34	24	of	of	ADP
ajis02-348	34	25	estimate	estimate	NOUN
ajis02-348	34	26	to	to	ADP
ajis02-348	34	27	actual	actual	ADJ
ajis02-348	34	28	value	value	NOUN
ajis02-348	34	29	.	.	PUNCT
ajis02-348	35	1	this	this	PRON
ajis02-348	35	2	is	be	AUX
ajis02-348	35	3	an	an	DET
ajis02-348	35	4	unbiased	unbiased	ADJ
ajis02-348	35	5	method	method	NOUN
ajis02-348	35	6	of	of	ADP
ajis02-348	35	7	finding	find	VERB
ajis02-348	35	8	parameters	parameter	NOUN
ajis02-348	35	9	in	in	ADP
ajis02-348	35	10	a	a	DET
ajis02-348	35	11	cost	cost	NOUN
ajis02-348	35	12	estimation	estimation	NOUN
ajis02-348	35	13	model	model	NOUN
ajis02-348	35	14	when	when	SCONJ
ajis02-348	35	15	the	the	DET
ajis02-348	35	16	functional	functional	ADJ
ajis02-348	35	17	form	form	NOUN
ajis02-348	35	18	of	of	ADP
ajis02-348	35	19	the	the	DET
ajis02-348	35	20	model	model	NOUN
ajis02-348	35	21	is	be	AUX
ajis02-348	35	22	known	know	VERB
ajis02-348	35	23	.	.	PUNCT
ajis02-348	36	1	the	the	DET
ajis02-348	36	2	third	third	ADJ
ajis02-348	36	3	part	part	NOUN
ajis02-348	36	4	assesses	assess	VERB
ajis02-348	36	5	the	the	DET
ajis02-348	36	6	proposed	propose	VERB
ajis02-348	36	7	measures	measure	NOUN
ajis02-348	36	8	and	and	CCONJ
ajis02-348	36	9	criterion	criterion	NOUN
ajis02-348	36	10	.	.	PUNCT
ajis02-348	37	1	a	a	DET
ajis02-348	37	2	data	datum	NOUN
ajis02-348	37	3	set	set	VERB
ajis02-348	37	4	from	from	ADP
ajis02-348	37	5	real	real	ADJ
ajis02-348	37	6	world	world	NOUN
ajis02-348	37	7	software	software	NOUN
ajis02-348	37	8	projects	project	NOUN
ajis02-348	37	9	was	be	AUX
ajis02-348	37	10	used	use	VERB
ajis02-348	37	11	to	to	PART
ajis02-348	37	12	examine	examine	VERB
ajis02-348	37	13	the	the	DET
ajis02-348	37	14	proposed	propose	VERB
ajis02-348	37	15	measures	measure	NOUN
ajis02-348	37	16	and	and	CCONJ
ajis02-348	37	17	the	the	DET
ajis02-348	37	18	regression	regression	NOUN
ajis02-348	37	19	criterion	criterion	NOUN
ajis02-348	37	20	so	so	SCONJ
ajis02-348	37	21	as	as	SCONJ
ajis02-348	37	22	to	to	PART
ajis02-348	37	23	demonstrate	demonstrate	VERB
ajis02-348	37	24	their	their	PRON
ajis02-348	37	25	advantages	advantage	NOUN
ajis02-348	37	26	over	over	ADP
ajis02-348	37	27	the	the	DET
ajis02-348	37	28	other	other	ADJ
ajis02-348	37	29	measures	measure	NOUN
ajis02-348	37	30	and	and	CCONJ
ajis02-348	37	31	criteria	criterion	NOUN
ajis02-348	37	32	currently	currently	ADV
ajis02-348	37	33	reported	report	VERB
ajis02-348	37	34	in	in	ADP
ajis02-348	37	35	the	the	DET
ajis02-348	37	36	literature	literature	NOUN
ajis02-348	37	37	.	.	PUNCT
ajis02-348	38	1	ajis	ajis	ADV
ajis02-348	38	2	vol	vol	VERB
ajis02-348	38	3	s	s	VERB
ajis02-348	38	4	no	no	DET
ajis02-348	38	5	1	1	NUM
ajis02-348	38	6	september	september	NOUN
ajis02-348	38	7	1997	1997	NUM
ajis02-348	38	8	31	31	NUM
ajis02-348	38	9	measures	measure	NOUN
ajis02-348	38	10	of	of	ADP
ajis02-348	38	11	accuracy	accuracy	NOUN
ajis02-348	38	12	accuracy	accuracy	NOUN
ajis02-348	38	13	is	be	AUX
ajis02-348	38	14	defined	define	VERB
ajis02-348	38	15	as	as	ADP
ajis02-348	38	16	the	the	DET
ajis02-348	38	17	measure	measure	NOUN
ajis02-348	38	18	of	of	ADP
ajis02-348	38	19	how	how	SCONJ
ajis02-348	38	20	close	close	ADJ
ajis02-348	38	21	a	a	DET
ajis02-348	38	22	result	result	NOUN
ajis02-348	38	23	is	be	AUX
ajis02-348	38	24	to	to	ADP
ajis02-348	38	25	its	its	PRON
ajis02-348	38	26	correct	correct	ADJ
ajis02-348	38	27	value	value	NOUN
ajis02-348	38	28	(	(	PUNCT
ajis02-348	38	29	deeson	deeson	NOUN
ajis02-348	38	30	,	,	PUNCT
ajis02-348	38	31	1991	1991	NUM
ajis02-348	38	32	)	)	PUNCT
ajis02-348	38	33	.	.	PUNCT
ajis02-348	39	1	there	there	PRON
ajis02-348	39	2	are	be	VERB
ajis02-348	39	3	two	two	NUM
ajis02-348	39	4	ways	way	NOUN
ajis02-348	39	5	to	to	PART
ajis02-348	39	6	compare	compare	VERB
ajis02-348	39	7	a	a	DET
ajis02-348	39	8	result	result	NOUN
ajis02-348	39	9	and	and	CCONJ
ajis02-348	39	10	its	its	PRON
ajis02-348	39	11	correct	correct	ADJ
ajis02-348	39	12	value	value	NOUN
ajis02-348	39	13	:	:	PUNCT
ajis02-348	39	14	their	their	PRON
ajis02-348	39	15	difference	difference	NOUN
ajis02-348	39	16	and	and	CCONJ
ajis02-348	39	17	their	their	PRON
ajis02-348	39	18	ratio	ratio	NOUN
ajis02-348	39	19	.	.	PUNCT
ajis02-348	40	1	let	let	VERB
ajis02-348	40	2	n	n	PRON
ajis02-348	40	3	be	be	AUX
ajis02-348	40	4	the	the	DET
ajis02-348	40	5	number	number	NOUN
ajis02-348	40	6	of	of	ADP
ajis02-348	40	7	projects	project	NOUN
ajis02-348	40	8	in	in	ADP
ajis02-348	40	9	a	a	DET
ajis02-348	40	10	data	data	NOUN
ajis02-348	40	11	set	set	VERB
ajis02-348	40	12	,	,	PUNCT
ajis02-348	40	13	acti	acti	AUX
ajis02-348	40	14	be	be	AUX
ajis02-348	40	15	the	the	DET
ajis02-348	40	16	/	/	SYM
ajis02-348	40	17	'	'	NOUN
ajis02-348	40	18	*	*	PUNCT
ajis02-348	40	19	(	(	PUNCT
ajis02-348	40	20	i=	i=	NOUN
ajis02-348	40	21	1,2,3	1,2,3	NUM
ajis02-348	40	22	,	,	PUNCT
ajis02-348	40	23	...	...	PUNCT
ajis02-348	40	24	,	,	PUNCT
ajis02-348	40	25	"	"	PUNCT
ajis02-348	40	26	)	)	PUNCT
ajis02-348	40	27	actual	actual	ADJ
ajis02-348	40	28	observed	observe	VERB
ajis02-348	40	29	value	value	NOUN
ajis02-348	40	30	and	and	CCONJ
ajis02-348	40	31	estj	estj	NOUN
ajis02-348	40	32	be	be	AUX
ajis02-348	40	33	the	the	DET
ajis02-348	40	34	corresponding	corresponding	ADJ
ajis02-348	40	35	estimated	estimate	VERB
ajis02-348	40	36	value	value	NOUN
ajis02-348	40	37	.	.	PUNCT
ajis02-348	41	1	the	the	DET
ajis02-348	41	2	difference	difference	NOUN
ajis02-348	41	3	measure	measure	NOUN
ajis02-348	41	4	of	of	ADP
ajis02-348	41	5	estimation	estimation	NOUN
ajis02-348	41	6	accuracy	accuracy	NOUN
ajis02-348	41	7	is	be	AUX
ajis02-348	41	8	based	base	VERB
ajis02-348	41	9	on	on	ADP
ajis02-348	41	10	the	the	DET
ajis02-348	41	11	difference	difference	NOUN
ajis02-348	41	12	between	between	ADP
ajis02-348	41	13	estimated	estimate	VERB
ajis02-348	41	14	value	value	NOUN
ajis02-348	41	15	and	and	CCONJ
ajis02-348	41	16	actual	actual	ADJ
ajis02-348	41	17	value	value	NOUN
ajis02-348	41	18	esti	esti	PROPN
ajis02-348	41	19	act	act	PROPN
ajis02-348	41	20	{	{	PUNCT
ajis02-348	41	21	(	(	PUNCT
ajis02-348	41	22	i	i	NOUN
ajis02-348	41	23	=	=	NOUN
ajis02-348	41	24	l,2,3,	l,2,3,	NUM
ajis02-348	41	25	...	...	PUNCT
ajis02-348	41	26	,n	,n	NOUN
ajis02-348	41	27	)	)	PUNCT
ajis02-348	41	28	the	the	DET
ajis02-348	41	29	ratio	ratio	NOUN
ajis02-348	41	30	measure	measure	NOUN
ajis02-348	41	31	of	of	ADP
ajis02-348	41	32	accuracy	accuracy	NOUN
ajis02-348	41	33	is	be	AUX
ajis02-348	41	34	based	base	VERB
ajis02-348	41	35	on	on	ADP
ajis02-348	41	36	the	the	DET
ajis02-348	41	37	ratio	ratio	NOUN
ajis02-348	41	38	of	of	ADP
ajis02-348	41	39	estimated	estimate	VERB
ajis02-348	41	40	value	value	NOUN
ajis02-348	41	41	to	to	ADP
ajis02-348	41	42	actual	actual	ADJ
ajis02-348	41	43	value	value	NOUN
ajis02-348	41	44	est	est	X
ajis02-348	41	45	{	{	PUNCT
ajis02-348	41	46	'	'	PUNCT
ajis02-348	41	47	(	(	PUNCT
ajis02-348	41	48	j	j	X
ajis02-348	41	49	=	=	NOUN
ajis02-348	41	50	l,2,3,	l,2,3,	NUM
ajis02-348	41	51	...	...	PUNCT
ajis02-348	41	52	,n	,n	NOUN
ajis02-348	41	53	)	)	PUNCT
ajis02-348	41	54	actf	actf	ADV
ajis02-348	41	55	in	in	ADP
ajis02-348	41	56	evaluating	evaluate	VERB
ajis02-348	41	57	the	the	DET
ajis02-348	41	58	accuracy	accuracy	NOUN
ajis02-348	41	59	of	of	ADP
ajis02-348	41	60	software	software	NOUN
ajis02-348	41	61	cost	cost	NOUN
ajis02-348	41	62	estimation	estimation	NOUN
ajis02-348	41	63	models	model	NOUN
ajis02-348	41	64	,	,	PUNCT
ajis02-348	41	65	both	both	DET
ajis02-348	41	66	difference	difference	NOUN
ajis02-348	41	67	and	and	CCONJ
ajis02-348	41	68	ratio	ratio	NOUN
ajis02-348	41	69	measures	measure	NOUN
ajis02-348	41	70	have	have	AUX
ajis02-348	41	71	been	be	AUX
ajis02-348	41	72	used	use	VERB
ajis02-348	41	73	.	.	PUNCT
ajis02-348	42	1	these	these	PRON
ajis02-348	42	2	are	be	AUX
ajis02-348	42	3	discussed	discuss	VERB
ajis02-348	42	4	below	below	ADV
ajis02-348	42	5	.	.	PUNCT
ajis02-348	43	1	difference	difference	NOUN
ajis02-348	43	2	measures	measure	NOUN
ajis02-348	43	3	of	of	ADP
ajis02-348	43	4	accuracy	accuracy	NOUN
ajis02-348	43	5	(	(	PUNCT
ajis02-348	43	6	1	1	NUM
ajis02-348	43	7	)	)	PUNCT
ajis02-348	43	8	mean	mean	NOUN
ajis02-348	43	9	of	of	ADP
ajis02-348	43	10	absolute	absolute	ADJ
ajis02-348	43	11	errors	error	NOUN
ajis02-348	43	12	(	(	PUNCT
ajis02-348	43	13	mae	mae	PROPN
ajis02-348	43	14	)	)	PUNCT
ajis02-348	43	15	i	i	PRON
ajis02-348	43	16	n	n	ADV
ajis02-348	43	17	mae=	mae=	PROPN
ajis02-348	43	18	—	—	PUNCT
ajis02-348	43	19	\\act	\\act	NOUN
ajis02-348	43	20	.	.	PUNCT
ajis02-348	43	21	-esti	-esti	NOUN
ajis02-348	43	22	"	"	PUNCT
ajis02-348	43	23	w	w	NOUN
ajis02-348	43	24	(	(	PUNCT
ajis02-348	43	25	2	2	NUM
ajis02-348	43	26	)	)	PUNCT
ajis02-348	43	27	root	root	NOUN
ajis02-348	43	28	mean	mean	NOUN
ajis02-348	43	29	of	of	ADP
ajis02-348	43	30	squares	square	NOUN
ajis02-348	43	31	of	of	ADP
ajis02-348	43	32	error	error	NOUN
ajis02-348	43	33	(	(	PUNCT
ajis02-348	43	34	rmse	rmse	NOUN
ajis02-348	43	35	)	)	PUNCT
ajis02-348	43	36	n	n	PROPN
ajis02-348	43	37	1=1	1=1	NUM
ajis02-348	43	38	(	(	PUNCT
ajis02-348	43	39	3	3	NUM
ajis02-348	43	40	)	)	PUNCT
ajis02-348	43	41	coefficient	coefficient	NOUN
ajis02-348	43	42	of	of	ADP
ajis02-348	43	43	determinant	determinant	ADJ
ajis02-348	43	44	(	(	PUNCT
ajis02-348	43	45	r2	r2	PROPN
ajis02-348	43	46	)	)	PUNCT
ajis02-348	44	1	_	_	NOUN
ajis02-348	44	2	1	1	NUM
ajis02-348	44	3	»	»	NOUN
ajis02-348	44	4	/•	/•	VERB
ajis02-348	44	5	_	_	PRON
ajis02-348	44	6	where	where	SCONJ
ajis02-348	44	7	act	act	NOUN
ajis02-348	44	8	=	=	PUNCT
ajis02-348	44	9	—	—	PUNCT
ajis02-348	44	10	y	y	PROPN
ajis02-348	44	11	acti	acti	VERB
ajis02-348	44	12	is	be	AUX
ajis02-348	44	13	the	the	DET
ajis02-348	44	14	mean	mean	NOUN
ajis02-348	44	15	of	of	ADP
ajis02-348	44	16	n	n	PRON
ajis02-348	44	17	actual	actual	ADJ
ajis02-348	44	18	observed	observed	ADJ
ajis02-348	44	19	values	value	NOUN
ajis02-348	44	20	.	.	PUNCT
ajis02-348	45	1	for	for	ADP
ajis02-348	45	2	a	a	DET
ajis02-348	45	3	given	give	VERB
ajis02-348	45	4	set	set	NOUN
ajis02-348	45	5	of	of	ADP
ajis02-348	45	6	data	datum	NOUN
ajis02-348	45	7	\	\	PROPN
ajis02-348	45	8	(	(	PUNCT
ajis02-348	45	9	act	act	PROPN
ajis02-348	45	10	,	,	PUNCT
ajis02-348	45	11	-act)2	-act)2	PROPN
ajis02-348	45	12	is	be	AUX
ajis02-348	45	13	a	a	DET
ajis02-348	45	14	n	n	NUM
ajis02-348	45	15	1=1	1=1	NUM
ajis02-348	45	16	1=1	1=1	NUM
ajis02-348	45	17	constant	constant	ADJ
ajis02-348	45	18	.	.	PUNCT
ajis02-348	46	1	therefore	therefore	ADV
ajis02-348	46	2	r	r	NOUN
ajis02-348	46	3	is	be	AUX
ajis02-348	46	4	a	a	DET
ajis02-348	46	5	distance	distance	NOUN
ajis02-348	46	6	measure	measure	NOUN
ajis02-348	46	7	.	.	PUNCT
ajis02-348	47	1	(	(	PUNCT
ajis02-348	47	2	4	4	X
ajis02-348	47	3	)	)	PUNCT
ajis02-348	47	4	mean	mean	NOUN
ajis02-348	47	5	of	of	ADP
ajis02-348	47	6	residues	residue	NOUN
ajis02-348	47	7	(	(	PUNCT
ajis02-348	47	8	mr	mr	PROPN
ajis02-348	47	9	)	)	PUNCT
ajis02-348	47	10	1	1	NUM
ajis02-348	48	1	"	"	PUNCT
ajis02-348	48	2	mr	mr	PROPN
ajis02-348	48	3	=	=	SYM
ajis02-348	48	4	̂	̂	PUNCT
ajis02-348	48	5	(	(	PUNCT
ajis02-348	48	6	estf	estf	NOUN
ajis02-348	48	7	~	~	PUNCT
ajis02-348	48	8	acti	acti	VERB
ajis02-348	48	9	)	)	PUNCT
ajis02-348	48	10	«	«	PUNCT
ajis02-348	48	11	1=1	1=1	NUM
ajis02-348	48	12	ratio	ratio	NOUN
ajis02-348	48	13	measures	measure	NOUN
ajis02-348	48	14	of	of	ADP
ajis02-348	48	15	accuracy	accuracy	NOUN
ajis02-348	48	16	(	(	PUNCT
ajis02-348	48	17	1	1	NUM
ajis02-348	48	18	)	)	PUNCT
ajis02-348	48	19	mean	mean	NOUN
ajis02-348	48	20	(	(	PUNCT
ajis02-348	48	21	or	or	CCONJ
ajis02-348	48	22	average	average	ADJ
ajis02-348	48	23	)	)	PUNCT
ajis02-348	48	24	of	of	ADP
ajis02-348	48	25	relative	relative	ADJ
ajis02-348	48	26	errors	error	NOUN
ajis02-348	48	27	(	(	PUNCT
ajis02-348	48	28	are	be	AUX
ajis02-348	48	29	)	)	PUNCT
ajis02-348	48	30	n	n	PROPN
ajis02-348	48	31	~	~	PUNCT
ajis02-348	48	32	act	act	NOUN
ajis02-348	48	33	,	,	PUNCT
ajis02-348	48	34	ajis	ajis	ADV
ajis02-348	48	35	vol	vol	VERB
ajis02-348	48	36	5	5	NUM
ajis02-348	48	37	no	no	DET
ajis02-348	48	38	1	1	NUM
ajis02-348	48	39	september	september	NOUN
ajis02-348	48	40	1997	1997	NUM
ajis02-348	48	41	32	32	NUM
ajis02-348	48	42	(	(	PUNCT
ajis02-348	48	43	2	2	NUM
ajis02-348	48	44	)	)	PUNCT
ajis02-348	48	45	mean	mean	NOUN
ajis02-348	48	46	of	of	ADP
ajis02-348	48	47	magnitude	magnitude	NOUN
ajis02-348	48	48	of	of	ADP
ajis02-348	48	49	relative	relative	ADJ
ajis02-348	48	50	errors	error	NOUN
ajis02-348	48	51	(	(	PUNCT
ajis02-348	48	52	mre	mre	NOUN
ajis02-348	48	53	)	)	PUNCT
ajis02-348	48	54	act	act	PROPN
ajis02-348	48	55	:	:	PUNCT
ajis02-348	48	56	est	est	X
ajis02-348	48	57	:	:	PUNCT
ajis02-348	48	58	act	act	VERB
ajis02-348	48	59	1est	1est	NUM
ajis02-348	48	60	.	.	PUNCT
ajis02-348	48	61	act	act	PROPN
ajis02-348	48	62	.	.	PUNCT
ajis02-348	49	1	(	(	PUNCT
ajis02-348	49	2	3	3	X
ajis02-348	49	3	)	)	PUNCT
ajis02-348	49	4	root	root	NOUN
ajis02-348	49	5	mean	mean	NOUN
ajis02-348	49	6	of	of	ADP
ajis02-348	49	7	squared	squared	ADJ
ajis02-348	49	8	relative	relative	ADJ
ajis02-348	49	9	errors	error	NOUN
ajis02-348	49	10	(	(	PUNCT
ajis02-348	49	11	rmsre	rmsre	NOUN
ajis02-348	49	12	)	)	PUNCT
ajis02-348	49	13	rmsre	rmsre	NOUN
ajis02-348	49	14	=	=	SYM
ajis02-348	49	15	\many	\many	ADJ
ajis02-348	49	16	measures	measure	NOUN
ajis02-348	49	17	are	be	AUX
ajis02-348	49	18	based	base	VERB
ajis02-348	49	19	on	on	ADP
ajis02-348	49	20	magnitude	magnitude	NOUN
ajis02-348	49	21	of	of	ADP
ajis02-348	49	22	relative	relative	ADJ
ajis02-348	49	23	errors	error	NOUN
ajis02-348	49	24	,	,	PUNCT
ajis02-348	49	25	mrei	mrei	NOUN
ajis02-348	49	26	mre	mre	PROPN
ajis02-348	49	27	,	,	PUNCT
ajis02-348	49	28	=	=	SYM
ajis02-348	49	29	act	act	PROPN
ajis02-348	49	30	.	.	PUNCT
ajis02-348	50	1	est	est	PROPN
ajis02-348	50	2	:	:	PUNCT
ajis02-348	50	3	act	act	VERB
ajis02-348	50	4	:	:	PUNCT
ajis02-348	50	5	mre	mre	PROPN
ajis02-348	50	6	is	be	AUX
ajis02-348	50	7	the	the	DET
ajis02-348	50	8	most	most	ADV
ajis02-348	50	9	widely	widely	ADV
ajis02-348	50	10	used	use	VERB
ajis02-348	50	11	measure	measure	NOUN
ajis02-348	50	12	in	in	ADP
ajis02-348	50	13	the	the	DET
ajis02-348	50	14	literature	literature	NOUN
ajis02-348	50	15	.	.	PUNCT
ajis02-348	51	1	however	however	ADV
ajis02-348	51	2	,	,	PUNCT
ajis02-348	51	3	it	it	PRON
ajis02-348	51	4	is	be	AUX
ajis02-348	51	5	influenced	influence	VERB
ajis02-348	51	6	by	by	ADP
ajis02-348	51	7	extreme	extreme	ADJ
ajis02-348	51	8	values	value	NOUN
ajis02-348	51	9	or	or	CCONJ
ajis02-348	51	10	outliers	outlier	NOUN
ajis02-348	51	11	.	.	PUNCT
ajis02-348	52	1	a	a	DET
ajis02-348	52	2	smaller	small	ADJ
ajis02-348	52	3	mre	mre	NOUN
ajis02-348	52	4	is	be	AUX
ajis02-348	52	5	not	not	PART
ajis02-348	52	6	always	always	ADV
ajis02-348	52	7	better	well	ADJ
ajis02-348	52	8	than	than	ADP
ajis02-348	52	9	a	a	DET
ajis02-348	52	10	larger	large	ADJ
ajis02-348	52	11	one	one	NOUN
ajis02-348	52	12	(	(	PUNCT
ajis02-348	52	13	j0rgensen	j0rgensen	PROPN
ajis02-348	52	14	,	,	PUNCT
ajis02-348	52	15	1994	1994	NUM
ajis02-348	52	16	)	)	PUNCT
ajis02-348	52	17	.	.	PUNCT
ajis02-348	53	1	to	to	PART
ajis02-348	53	2	address	address	VERB
ajis02-348	53	3	this	this	DET
ajis02-348	53	4	problem	problem	NOUN
ajis02-348	53	5	,	,	PUNCT
ajis02-348	53	6	some	some	DET
ajis02-348	53	7	single	single	ADJ
ajis02-348	53	8	values	value	NOUN
ajis02-348	53	9	of	of	ADP
ajis02-348	53	10	mrei	mrei	NOUN
ajis02-348	53	11	(	(	PUNCT
ajis02-348	53	12	i=\,2,3,	i=\,2,3,	NOUN
ajis02-348	53	13	...	...	PUNCT
ajis02-348	53	14	,ri	,ri	PUNCT
ajis02-348	53	15	)	)	PUNCT
ajis02-348	53	16	have	have	AUX
ajis02-348	53	17	been	be	AUX
ajis02-348	53	18	used	use	VERB
ajis02-348	53	19	to	to	PART
ajis02-348	53	20	measure	measure	VERB
ajis02-348	53	21	the	the	DET
ajis02-348	53	22	estimation	estimation	NOUN
ajis02-348	53	23	accuracy	accuracy	NOUN
ajis02-348	53	24	.	.	PUNCT
ajis02-348	54	1	they	they	PRON
ajis02-348	54	2	are	be	AUX
ajis02-348	54	3	described	describe	VERB
ajis02-348	54	4	below	below	ADV
ajis02-348	54	5	.	.	PUNCT
ajis02-348	55	1	(	(	PUNCT
ajis02-348	55	2	4	4	X
ajis02-348	55	3	)	)	PUNCT
ajis02-348	55	4	prediction	prediction	NOUN
ajis02-348	55	5	at	at	ADP
ajis02-348	55	6	level	level	NOUN
ajis02-348	55	7	/	/	SYM
ajis02-348	55	8	(	(	PUNCT
ajis02-348	55	9	pred(/	pred(/	NOUN
ajis02-348	55	10	)	)	PUNCT
ajis02-348	55	11	)	)	PUNCT
ajis02-348	56	1	conte	conte	PROPN
ajis02-348	56	2	el	el	PROPN
ajis02-348	56	3	al	al	PROPN
ajis02-348	56	4	(	(	PUNCT
ajis02-348	56	5	1986	1986	NUM
ajis02-348	56	6	)	)	PUNCT
ajis02-348	56	7	put	put	VERB
ajis02-348	56	8	forward	forward	ADV
ajis02-348	56	9	a	a	DET
ajis02-348	56	10	single	single	ADJ
ajis02-348	56	11	value	value	NOUN
ajis02-348	56	12	measure	measure	NOUN
ajis02-348	56	13	,	,	PUNCT
ajis02-348	56	14	prediction	prediction	NOUN
ajis02-348	56	15	at	at	ADP
ajis02-348	56	16	level	level	NOUN
ajis02-348	56	17	/	/	SYM
ajis02-348	56	18	pred(l	pred(l	NOUN
ajis02-348	56	19	)	)	PUNCT
ajis02-348	56	20	=	=	SYM
ajis02-348	56	21	—	—	PUNCT
ajis02-348	56	22	n	n	CCONJ
ajis02-348	56	23	where	where	SCONJ
ajis02-348	56	24	k	k	PROPN
ajis02-348	56	25	is	be	AUX
ajis02-348	56	26	the	the	DET
ajis02-348	56	27	number	number	NOUN
ajis02-348	56	28	of	of	ADP
ajis02-348	56	29	projects	project	NOUN
ajis02-348	56	30	in	in	ADP
ajis02-348	56	31	a	a	DET
ajis02-348	56	32	set	set	NOUN
ajis02-348	56	33	of	of	ADP
ajis02-348	56	34	n	n	NOUN
ajis02-348	56	35	projects	project	NOUN
ajis02-348	56	36	whose	whose	DET
ajis02-348	56	37	mre	mre	NOUN
ajis02-348	56	38	<	<	X
ajis02-348	56	39	i.	i.	PROPN
ajis02-348	56	40	they	they	PRON
ajis02-348	56	41	suggested	suggest	VERB
ajis02-348	56	42	that	that	SCONJ
ajis02-348	56	43	an	an	DET
ajis02-348	56	44	acceptable	acceptable	ADJ
ajis02-348	56	45	accuracy	accuracy	NOUN
ajis02-348	56	46	for	for	ADP
ajis02-348	56	47	a	a	DET
ajis02-348	56	48	model	model	NOUN
ajis02-348	56	49	is	be	AUX
ajis02-348	56	50	pred(0.25	pred(0.25	NOUN
ajis02-348	56	51	)	)	PUNCT
ajis02-348	56	52	>	>	X
ajis02-348	57	1	0.75	0.75	NUM
ajis02-348	57	2	,	,	PUNCT
ajis02-348	57	3	which	which	PRON
ajis02-348	57	4	is	be	AUX
ajis02-348	57	5	seldom	seldom	ADV
ajis02-348	57	6	reached	reach	VERB
ajis02-348	57	7	in	in	ADP
ajis02-348	57	8	reality	reality	NOUN
ajis02-348	57	9	.	.	PUNCT
ajis02-348	58	1	(	(	PUNCT
ajis02-348	58	2	5	5	NUM
ajis02-348	58	3	)	)	PUNCT
ajis02-348	58	4	third	third	ADJ
ajis02-348	58	5	quartile	quartile	NOUN
ajis02-348	58	6	of	of	ADP
ajis02-348	58	7	mre	mre	PROPN
ajis02-348	58	8	(	(	PUNCT
ajis02-348	58	9	q3	q3	PROPN
ajis02-348	58	10	)	)	PUNCT
ajis02-348	58	11	conte	conte	PROPN
ajis02-348	58	12	et	et	PROPN
ajis02-348	58	13	al	al	PROPN
ajis02-348	58	14	's	's	PART
ajis02-348	58	15	standard	standard	NOUN
ajis02-348	58	16	can	can	AUX
ajis02-348	58	17	be	be	AUX
ajis02-348	58	18	expressed	express	VERB
ajis02-348	58	19	in	in	ADP
ajis02-348	58	20	another	another	DET
ajis02-348	58	21	way	way	NOUN
ajis02-348	58	22	:	:	PUNCT
ajis02-348	58	23	75	75	NUM
ajis02-348	58	24	%	%	NOUN
ajis02-348	58	25	of	of	ADP
ajis02-348	58	26	the	the	DET
ajis02-348	58	27	mre	mre	PROPN
ajis02-348	58	28	's	's	PART
ajis02-348	58	29	are	be	AUX
ajis02-348	58	30	less	less	ADJ
ajis02-348	58	31	than	than	ADP
ajis02-348	58	32	or	or	CCONJ
ajis02-348	58	33	equal	equal	ADJ
ajis02-348	58	34	to	to	ADP
ajis02-348	58	35	0.25	0.25	NUM
ajis02-348	58	36	.	.	PUNCT
ajis02-348	59	1	in	in	ADP
ajis02-348	59	2	terms	term	NOUN
ajis02-348	59	3	of	of	ADP
ajis02-348	59	4	quartiles	quartile	NOUN
ajis02-348	59	5	,	,	PUNCT
ajis02-348	59	6	the	the	DET
ajis02-348	59	7	third	third	ADJ
ajis02-348	59	8	quartile	quartile	NOUN
ajis02-348	59	9	,	,	PUNCT
ajis02-348	59	10	q3	q3	PROPN
ajis02-348	59	11	,	,	PUNCT
ajis02-348	59	12	of	of	ADP
ajis02-348	59	13	the	the	DET
ajis02-348	59	14	mre	mre	PROPN
ajis02-348	59	15	's	's	PART
ajis02-348	59	16	is	be	AUX
ajis02-348	59	17	less	less	ADJ
ajis02-348	59	18	than	than	ADP
ajis02-348	59	19	or	or	CCONJ
ajis02-348	59	20	equal	equal	ADJ
ajis02-348	59	21	to	to	ADP
ajis02-348	59	22	0.25	0.25	NUM
ajis02-348	59	23	,	,	PUNCT
ajis02-348	59	24	i.e.	i.e.	X
ajis02-348	59	25	q3	q3	NOUN
ajis02-348	59	26	<	<	X
ajis02-348	59	27	0.25	0.25	NUM
ajis02-348	59	28	.	.	PUNCT
ajis02-348	60	1	as	as	SCONJ
ajis02-348	60	2	mentioned	mention	VERB
ajis02-348	60	3	above	above	ADV
ajis02-348	60	4	,	,	PUNCT
ajis02-348	60	5	this	this	PRON
ajis02-348	60	6	is	be	AUX
ajis02-348	60	7	too	too	ADV
ajis02-348	60	8	high	high	ADJ
ajis02-348	60	9	a	a	DET
ajis02-348	60	10	standard	standard	NOUN
ajis02-348	60	11	at	at	ADP
ajis02-348	60	12	present	present	NOUN
ajis02-348	60	13	,	,	PUNCT
ajis02-348	60	14	but	but	CCONJ
ajis02-348	60	15	it	it	PRON
ajis02-348	60	16	should	should	AUX
ajis02-348	60	17	be	be	AUX
ajis02-348	60	18	a	a	DET
ajis02-348	60	19	goal	goal	NOUN
ajis02-348	60	20	to	to	PART
ajis02-348	60	21	pursue	pursue	VERB
ajis02-348	60	22	.	.	PUNCT
ajis02-348	61	1	the	the	PRON
ajis02-348	61	2	smaller	small	ADJ
ajis02-348	61	3	the	the	DET
ajis02-348	61	4	<	<	X
ajis02-348	61	5	23	23	NUM
ajis02-348	61	6	,	,	PUNCT
ajis02-348	61	7	the	the	DET
ajis02-348	61	8	more	more	ADV
ajis02-348	61	9	accurate	accurate	ADJ
ajis02-348	61	10	the	the	DET
ajis02-348	61	11	estimation	estimation	NOUN
ajis02-348	61	12	.	.	PUNCT
ajis02-348	62	1	both	both	DET
ajis02-348	62	2	qj	qj	PROPN
ajis02-348	62	3	anc	anc	PROPN
ajis02-348	62	4	'	'	PART
ajis02-348	62	5	pred(0.25	pred(0.25	NOUN
ajis02-348	62	6	)	)	PUNCT
ajis02-348	62	7	avoid	avoid	VERB
ajis02-348	62	8	considering	consider	VERB
ajis02-348	62	9	some	some	DET
ajis02-348	62	10	extremely	extremely	ADV
ajis02-348	62	11	poor	poor	ADJ
ajis02-348	62	12	predicted	predict	VERB
ajis02-348	62	13	values	value	NOUN
ajis02-348	62	14	.	.	PUNCT
ajis02-348	63	1	so	so	ADV
ajis02-348	63	2	they	they	PRON
ajis02-348	63	3	eliminate	eliminate	VERB
ajis02-348	63	4	the	the	DET
ajis02-348	63	5	influence	influence	NOUN
ajis02-348	63	6	of	of	ADP
ajis02-348	63	7	extreme	extreme	ADJ
ajis02-348	63	8	values	value	NOUN
ajis02-348	63	9	.	.	PUNCT
ajis02-348	64	1	(	(	PUNCT
ajis02-348	64	2	6	6	X
ajis02-348	64	3	)	)	PUNCT
ajis02-348	64	4	other	other	ADJ
ajis02-348	64	5	measures	measure	NOUN
ajis02-348	64	6	the	the	DET
ajis02-348	64	7	median	median	NOUN
ajis02-348	64	8	of	of	ADP
ajis02-348	64	9	mre	mre	PROPN
ajis02-348	64	10	's	's	PART
ajis02-348	64	11	(	(	PUNCT
ajis02-348	64	12	q2	q2	NOUN
ajis02-348	64	13	,	,	PUNCT
ajis02-348	64	14	second	second	ADJ
ajis02-348	64	15	quartile	quartile	NOUN
ajis02-348	64	16	)	)	PUNCT
ajis02-348	64	17	has	have	AUX
ajis02-348	64	18	also	also	ADV
ajis02-348	64	19	been	be	AUX
ajis02-348	64	20	used	use	VERB
ajis02-348	64	21	as	as	ADP
ajis02-348	64	22	one	one	NUM
ajis02-348	64	23	of	of	ADP
ajis02-348	64	24	the	the	DET
ajis02-348	64	25	measures	measure	NOUN
ajis02-348	64	26	of	of	ADP
ajis02-348	64	27	accuracy	accuracy	NOUN
ajis02-348	64	28	(	(	PUNCT
ajis02-348	64	29	j0rgensen	j0rgensen	PROPN
ajis02-348	64	30	,	,	PUNCT
ajis02-348	64	31	1994	1994	NUM
ajis02-348	64	32	)	)	PUNCT
ajis02-348	64	33	.	.	PUNCT
ajis02-348	65	1	it	it	PRON
ajis02-348	65	2	is	be	AUX
ajis02-348	65	3	only	only	ADV
ajis02-348	65	4	a	a	DET
ajis02-348	65	5	middle	middle	ADJ
ajis02-348	65	6	number	number	NOUN
ajis02-348	65	7	among	among	ADP
ajis02-348	65	8	the	the	DET
ajis02-348	65	9	ordered	ordered	ADJ
ajis02-348	65	10	mre	mre	PROPN
ajis02-348	65	11	's	's	PART
ajis02-348	65	12	,	,	PUNCT
ajis02-348	65	13	so	so	CCONJ
ajis02-348	65	14	it	it	PRON
ajis02-348	65	15	hardly	hardly	ADV
ajis02-348	65	16	presents	present	VERB
ajis02-348	65	17	an	an	DET
ajis02-348	65	18	average	average	NOUN
ajis02-348	65	19	of	of	ADP
ajis02-348	65	20	all	all	DET
ajis02-348	65	21	the	the	DET
ajis02-348	65	22	mre	mre	NOUN
ajis02-348	65	23	's	's	PART
ajis02-348	65	24	.	.	PUNCT
ajis02-348	66	1	miyazaki	miyazaki	PROPN
ajis02-348	66	2	era/	era/	PROPN
ajis02-348	66	3	(	(	PUNCT
ajis02-348	66	4	1991	1991	NUM
ajis02-348	66	5	)	)	PUNCT
ajis02-348	66	6	believed	believe	VERB
ajis02-348	66	7	that	that	SCONJ
ajis02-348	66	8	mre	mre	PROPN
ajis02-348	66	9	based	base	VERB
ajis02-348	66	10	measures	measure	NOUN
ajis02-348	66	11	favour	favour	VERB
ajis02-348	66	12	underestimation	underestimation	NOUN
ajis02-348	66	13	.	.	PUNCT
ajis02-348	67	1	(	(	PUNCT
ajis02-348	67	2	this	this	DET
ajis02-348	67	3	problem	problem	NOUN
ajis02-348	67	4	will	will	AUX
ajis02-348	67	5	be	be	AUX
ajis02-348	67	6	discussed	discuss	VERB
ajis02-348	67	7	in	in	ADP
ajis02-348	67	8	detail	detail	NOUN
ajis02-348	67	9	later	later	ADV
ajis02-348	67	10	.	.	PUNCT
ajis02-348	67	11	)	)	PUNCT
ajis02-348	68	1	to	to	PART
ajis02-348	68	2	address	address	VERB
ajis02-348	68	3	this	this	DET
ajis02-348	68	4	problem	problem	NOUN
ajis02-348	68	5	,	,	PUNCT
ajis02-348	68	6	they	they	PRON
ajis02-348	68	7	defined	define	VERB
ajis02-348	68	8	the	the	DET
ajis02-348	68	9	relative	relative	ADJ
ajis02-348	68	10	error	error	NOUN
ajis02-348	68	11	in	in	ADP
ajis02-348	68	12	another	another	DET
ajis02-348	68	13	way	way	NOUN
ajis02-348	68	14	:	:	PUNCT
ajis02-348	68	15	est	est	X
ajis02-348	68	16	—	—	PUNCT
ajis02-348	68	17	act	act	VERB
ajis02-348	68	18	r.	r.	PROPN
ajis02-348	68	19	=	=	PROPN
ajis02-348	68	20	•	•	NOUN
ajis02-348	68	21	r	r	NOUN
ajis02-348	68	22	=	=	SYM
ajis02-348	68	23	•	•	PROPN
ajis02-348	68	24	esti	esti	PROPN
ajis02-348	68	25	act	act	PROPN
ajis02-348	68	26	(	(	PUNCT
ajis02-348	68	27	est	est	PROPN
ajis02-348	68	28	:	:	PUNCT
ajis02-348	68	29	{	{	PUNCT
ajis02-348	68	30	acti	acti	VERB
ajis02-348	68	31	>	>	X
ajis02-348	68	32	0	0	NUM
ajis02-348	68	33	)	)	PUNCT
ajis02-348	68	34	(	(	PUNCT
ajis02-348	68	35	est	est	X
ajis02-348	68	36	.	.	PUNCT
ajis02-348	68	37	acti	acti	VERB
ajis02-348	68	38	<	<	X
ajis02-348	68	39	0	0	NUM
ajis02-348	68	40	)	)	PUNCT
ajis02-348	68	41	ajis	ajis	ADV
ajis02-348	68	42	vol	vol	VERB
ajis02-348	68	43	5	5	NUM
ajis02-348	68	44	no	no	DET
ajis02-348	68	45	1	1	NUM
ajis02-348	68	46	september	september	NOUN
ajis02-348	68	47	1997	1997	NUM
ajis02-348	68	48	33	33	NUM
ajis02-348	68	49	they	they	PRON
ajis02-348	68	50	argued	argue	VERB
ajis02-348	68	51	that	that	SCONJ
ajis02-348	68	52	by	by	ADP
ajis02-348	68	53	using	use	VERB
ajis02-348	68	54	/	/	PUNCT
ajis02-348	68	55	"	"	PUNCT
ajis02-348	68	56	,	,	PUNCT
ajis02-348	68	57	,	,	PUNCT
ajis02-348	68	58	a	a	DET
ajis02-348	68	59	balanced	balanced	ADJ
ajis02-348	68	60	evaluation	evaluation	NOUN
ajis02-348	68	61	is	be	AUX
ajis02-348	68	62	obtained	obtain	VERB
ajis02-348	68	63	for	for	ADP
ajis02-348	68	64	both	both	CCONJ
ajis02-348	68	65	underestimated	underestimate	VERB
ajis02-348	68	66	and	and	CCONJ
ajis02-348	68	67	overestimated	overestimated	ADJ
ajis02-348	68	68	cases	case	NOUN
ajis02-348	68	69	.	.	PUNCT
ajis02-348	69	1	with	with	ADP
ajis02-348	69	2	the	the	DET
ajis02-348	69	3	above	above	ADJ
ajis02-348	69	4	definition	definition	NOUN
ajis02-348	69	5	of	of	ADP
ajis02-348	69	6	relative	relative	ADJ
ajis02-348	69	7	error	error	NOUN
ajis02-348	69	8	,	,	PUNCT
ajis02-348	69	9	mre	mre	NOUN
ajis02-348	69	10	,	,	PUNCT
ajis02-348	69	11	rmsre	rmsre	NOUN
ajis02-348	69	12	or	or	CCONJ
ajis02-348	69	13	pred(0.25	pred(0.25	NOUN
ajis02-348	69	14	)	)	PUNCT
ajis02-348	69	15	can	can	AUX
ajis02-348	69	16	be	be	AUX
ajis02-348	69	17	used	use	VERB
ajis02-348	69	18	for	for	ADP
ajis02-348	69	19	accuracy	accuracy	NOUN
ajis02-348	69	20	evaluation	evaluation	NOUN
ajis02-348	69	21	(	(	PUNCT
ajis02-348	69	22	miyazaki	miyazaki	PROPN
ajis02-348	69	23	et	et	PROPN
ajis02-348	69	24	al	al	PROPN
ajis02-348	69	25	,	,	PUNCT
ajis02-348	69	26	1991	1991	NUM
ajis02-348	69	27	)	)	PUNCT
ajis02-348	69	28	.	.	PUNCT
ajis02-348	70	1	fenton	fenton	PROPN
ajis02-348	70	2	(	(	PUNCT
ajis02-348	70	3	1994	1994	NUM
ajis02-348	70	4	)	)	PUNCT
ajis02-348	70	5	argued	argue	VERB
ajis02-348	70	6	that	that	SCONJ
ajis02-348	70	7	good	good	ADJ
ajis02-348	70	8	measurement	measurement	NOUN
ajis02-348	70	9	should	should	AUX
ajis02-348	70	10	be	be	AUX
ajis02-348	70	11	meaningful	meaningful	ADJ
ajis02-348	70	12	.	.	PUNCT
ajis02-348	71	1	in	in	ADP
ajis02-348	71	2	practice	practice	NOUN
ajis02-348	71	3	,	,	PUNCT
ajis02-348	71	4	it	it	PRON
ajis02-348	71	5	is	be	AUX
ajis02-348	71	6	generally	generally	ADV
ajis02-348	71	7	accepted	accept	VERB
ajis02-348	71	8	to	to	PART
ajis02-348	71	9	express	express	VERB
ajis02-348	71	10	relative	relative	ADJ
ajis02-348	71	11	error	error	NOUN
ajis02-348	71	12	by	by	ADP
ajis02-348	71	13	comparing	compare	VERB
ajis02-348	71	14	error	error	NOUN
ajis02-348	71	15	with	with	ADP
ajis02-348	71	16	actual	actual	ADJ
ajis02-348	71	17	observed	observed	ADJ
ajis02-348	71	18	value	value	NOUN
ajis02-348	71	19	.	.	PUNCT
ajis02-348	72	1	with	with	ADP
ajis02-348	72	2	this	this	DET
ajis02-348	72	3	commonly	commonly	ADV
ajis02-348	72	4	accepted	accept	VERB
ajis02-348	72	5	idea	idea	NOUN
ajis02-348	72	6	of	of	ADP
ajis02-348	72	7	relative	relative	ADJ
ajis02-348	72	8	error	error	NOUN
ajis02-348	72	9	,	,	PUNCT
ajis02-348	72	10	if	if	SCONJ
ajis02-348	72	11	the	the	DET
ajis02-348	72	12	relative	relative	ADJ
ajis02-348	72	13	errors	error	NOUN
ajis02-348	72	14	are	be	AUX
ajis02-348	72	15	0.40	0.40	NUM
ajis02-348	72	16	,	,	PUNCT
ajis02-348	72	17	0.50	0.50	NUM
ajis02-348	72	18	,	,	PUNCT
ajis02-348	72	19	0.60	0.60	NUM
ajis02-348	72	20	and	and	CCONJ
ajis02-348	72	21	0.70	0.70	NUM
ajis02-348	72	22	in	in	ADP
ajis02-348	72	23	the	the	DET
ajis02-348	72	24	case	case	NOUN
ajis02-348	72	25	of	of	ADP
ajis02-348	72	26	underestimation	underestimation	NOUN
ajis02-348	72	27	,	,	PUNCT
ajis02-348	72	28	the	the	DET
ajis02-348	72	29	corresponding	correspond	VERB
ajis02-348	72	30	r	r	NOUN
ajis02-348	72	31	values	value	NOUN
ajis02-348	72	32	are	be	AUX
ajis02-348	72	33	0.67	0.67	NUM
ajis02-348	72	34	,	,	PUNCT
ajis02-348	72	35	1.00	1.00	NUM
ajis02-348	72	36	,	,	PUNCT
ajis02-348	72	37	1.50	1.50	NUM
ajis02-348	72	38	and	and	CCONJ
ajis02-348	72	39	2.33	2.33	NUM
ajis02-348	72	40	.	.	PUNCT
ajis02-348	73	1	in	in	ADP
ajis02-348	73	2	an	an	DET
ajis02-348	73	3	extreme	extreme	ADJ
ajis02-348	73	4	case	case	NOUN
ajis02-348	73	5	,	,	PUNCT
ajis02-348	73	6	if	if	SCONJ
ajis02-348	73	7	the	the	DET
ajis02-348	73	8	relative	relative	ADJ
ajis02-348	73	9	error	error	NOUN
ajis02-348	73	10	is	be	AUX
ajis02-348	73	11	0.95	0.95	NUM
ajis02-348	73	12	,	,	PUNCT
ajis02-348	73	13	which	which	PRON
ajis02-348	73	14	is	be	AUX
ajis02-348	73	15	not	not	PART
ajis02-348	73	16	unusual	unusual	ADJ
ajis02-348	73	17	,	,	PUNCT
ajis02-348	73	18	the	the	DET
ajis02-348	73	19	r	r	NOUN
ajis02-348	73	20	value	value	NOUN
ajis02-348	73	21	can	can	AUX
ajis02-348	73	22	be	be	AUX
ajis02-348	73	23	as	as	ADV
ajis02-348	73	24	high	high	ADJ
ajis02-348	73	25	as	as	ADP
ajis02-348	73	26	19.00	19.00	NUM
ajis02-348	73	27	.	.	PUNCT
ajis02-348	74	1	the	the	DET
ajis02-348	74	2	variance	variance	NOUN
ajis02-348	74	3	of	of	ADP
ajis02-348	74	4	r	r	NOUN
ajis02-348	74	5	(	(	PUNCT
ajis02-348	74	6	must	must	AUX
ajis02-348	74	7	be	be	AUX
ajis02-348	74	8	large	large	ADJ
ajis02-348	74	9	and	and	CCONJ
ajis02-348	74	10	the	the	DET
ajis02-348	74	11	outliers	outlier	NOUN
ajis02-348	74	12	take	take	VERB
ajis02-348	74	13	an	an	DET
ajis02-348	74	14	important	important	ADJ
ajis02-348	74	15	role	role	NOUN
ajis02-348	74	16	in	in	ADP
ajis02-348	74	17	mre	mre	PROPN
ajis02-348	74	18	.	.	PUNCT
ajis02-348	75	1	the	the	DET
ajis02-348	75	2	large	large	ADJ
ajis02-348	75	3	variance	variance	NOUN
ajis02-348	75	4	will	will	AUX
ajis02-348	75	5	also	also	ADV
ajis02-348	75	6	make	make	VERB
ajis02-348	75	7	rmsre	rmsre	NOUN
ajis02-348	75	8	and	and	CCONJ
ajis02-348	75	9	pred(0.25	pred(0.25	NOUN
ajis02-348	75	10	)	)	PUNCT
ajis02-348	75	11	values	value	NOUN
ajis02-348	75	12	quite	quite	ADV
ajis02-348	75	13	different	different	ADJ
ajis02-348	75	14	from	from	ADP
ajis02-348	75	15	those	those	PRON
ajis02-348	75	16	commonly	commonly	ADV
ajis02-348	75	17	accepted	accept	VERB
ajis02-348	75	18	.	.	PUNCT
ajis02-348	76	1	moreover	moreover	ADV
ajis02-348	76	2	this	this	DET
ajis02-348	76	3	measure	measure	NOUN
ajis02-348	76	4	penalises	penalise	NOUN
ajis02-348	76	5	underestimation	underestimation	VERB
ajis02-348	76	6	too	too	ADV
ajis02-348	76	7	much	much	ADJ
ajis02-348	76	8	.	.	PUNCT
ajis02-348	77	1	jenson	jenson	PROPN
ajis02-348	77	2	and	and	CCONJ
ajis02-348	77	3	hartley	hartley	PROPN
ajis02-348	77	4	(	(	PUNCT
ajis02-348	77	5	1991	1991	NUM
ajis02-348	77	6	)	)	PUNCT
ajis02-348	77	7	proposed	propose	VERB
ajis02-348	77	8	the	the	DET
ajis02-348	77	9	weighted	weighted	ADJ
ajis02-348	77	10	mre	mre	PROPN
ajis02-348	77	11	(	(	PUNCT
ajis02-348	77	12	wmre	wmre	PROPN
ajis02-348	77	13	)	)	PUNCT
ajis02-348	77	14	.	.	PUNCT
ajis02-348	78	1	they	they	PRON
ajis02-348	78	2	believe	believe	VERB
ajis02-348	78	3	that	that	SCONJ
ajis02-348	78	4	one	one	NUM
ajis02-348	78	5	weakness	weakness	NOUN
ajis02-348	78	6	of	of	ADP
ajis02-348	78	7	mre	mre	PROPN
ajis02-348	78	8	is	be	AUX
ajis02-348	78	9	that	that	SCONJ
ajis02-348	78	10	the	the	DET
ajis02-348	78	11	relative	relative	ADJ
ajis02-348	78	12	size	size	NOUN
ajis02-348	78	13	of	of	ADP
ajis02-348	78	14	each	each	DET
ajis02-348	78	15	individual	individual	ADJ
ajis02-348	78	16	observation	observation	NOUN
ajis02-348	78	17	is	be	AUX
ajis02-348	78	18	not	not	PART
ajis02-348	78	19	recognised	recognise	VERB
ajis02-348	78	20	.	.	PUNCT
ajis02-348	79	1	wmre	wmre	PROPN
ajis02-348	79	2	is	be	AUX
ajis02-348	79	3	calculated	calculate	VERB
ajis02-348	79	4	such	such	ADJ
ajis02-348	79	5	that	that	SCONJ
ajis02-348	79	6	each	each	DET
ajis02-348	79	7	observation	observation	NOUN
ajis02-348	79	8	's	's	PART
ajis02-348	79	9	relative	relative	ADJ
ajis02-348	79	10	error	error	NOUN
ajis02-348	79	11	is	be	AUX
ajis02-348	79	12	weighted	weight	VERB
ajis02-348	79	13	by	by	ADP
ajis02-348	79	14	the	the	DET
ajis02-348	79	15	ratio	ratio	NOUN
ajis02-348	79	16	of	of	ADP
ajis02-348	79	17	the	the	DET
ajis02-348	79	18	observed	observe	VERB
ajis02-348	79	19	value	value	NOUN
ajis02-348	79	20	to	to	ADP
ajis02-348	79	21	the	the	DET
ajis02-348	79	22	average	average	NOUN
ajis02-348	79	23	of	of	ADP
ajis02-348	79	24	the	the	DET
ajis02-348	79	25	observed	observe	VERB
ajis02-348	79	26	values	value	NOUN
ajis02-348	79	27	.	.	PUNCT
ajis02-348	80	1	they	they	PRON
ajis02-348	80	2	argued	argue	VERB
ajis02-348	80	3	that	that	SCONJ
ajis02-348	80	4	by	by	ADP
ajis02-348	80	5	weighting	weight	VERB
ajis02-348	80	6	larger	large	ADJ
ajis02-348	80	7	projects	project	NOUN
ajis02-348	80	8	more	more	ADV
ajis02-348	80	9	heavily	heavily	ADV
ajis02-348	80	10	than	than	ADP
ajis02-348	80	11	smaller	small	ADJ
ajis02-348	80	12	ones	one	NOUN
ajis02-348	80	13	,	,	PUNCT
ajis02-348	80	14	the	the	DET
ajis02-348	80	15	wmre	wmre	NOUN
ajis02-348	80	16	recognises	recognise	VERB
ajis02-348	80	17	the	the	DET
ajis02-348	80	18	greater	great	ADJ
ajis02-348	80	19	cost	cost	NOUN
ajis02-348	80	20	associated	associate	VERB
ajis02-348	80	21	with	with	ADP
ajis02-348	80	22	estimation	estimation	NOUN
ajis02-348	80	23	errors	error	NOUN
ajis02-348	80	24	on	on	ADP
ajis02-348	80	25	larger	large	ADJ
ajis02-348	80	26	projects	project	NOUN
ajis02-348	80	27	.	.	PUNCT
ajis02-348	81	1	however	however	ADV
ajis02-348	81	2	,	,	PUNCT
ajis02-348	81	3	on	on	ADP
ajis02-348	81	4	closer	close	ADJ
ajis02-348	81	5	examination	examination	NOUN
ajis02-348	81	6	,	,	PUNCT
ajis02-348	81	7	we	we	PRON
ajis02-348	81	8	can	can	AUX
ajis02-348	81	9	prove	prove	VERB
ajis02-348	81	10	that	that	SCONJ
ajis02-348	81	11	their	their	PRON
ajis02-348	81	12	argument	argument	NOUN
ajis02-348	81	13	is	be	AUX
ajis02-348	81	14	untenable	untenable	ADJ
ajis02-348	81	15	since	since	SCONJ
ajis02-348	81	16	act	act	PROPN
ajis02-348	81	17	w*rr	w*rr	PROPN
ajis02-348	81	18	act	act	VERB
ajis02-348	81	19	act,-est	act,-est	ADV
ajis02-348	81	20	:	:	PUNCT
ajis02-348	81	21	act	act	NOUN
ajis02-348	81	22	;	;	PUNCT
ajis02-348	81	23	1	1	NUM
ajis02-348	81	24	nact^	nact^	NOUN
ajis02-348	81	25	act	act	NOUN
ajis02-348	81	26	,	,	PUNCT
ajis02-348	81	27	est	est	PROPN
ajis02-348	81	28	:	:	PUNCT
ajis02-348	81	29	in	in	ADP
ajis02-348	81	30	this	this	DET
ajis02-348	81	31	expression	expression	NOUN
ajis02-348	81	32	,	,	PUNCT
ajis02-348	81	33	it	it	PRON
ajis02-348	81	34	does	do	AUX
ajis02-348	81	35	not	not	PART
ajis02-348	81	36	appear	appear	VERB
ajis02-348	81	37	that	that	SCONJ
ajis02-348	81	38	larger	large	ADJ
ajis02-348	81	39	projects	project	NOUN
ajis02-348	81	40	are	be	AUX
ajis02-348	81	41	weighted	weight	VERB
ajis02-348	81	42	more	more	ADV
ajis02-348	81	43	heavily	heavily	ADV
ajis02-348	81	44	than	than	ADP
ajis02-348	81	45	smaller	small	ADJ
ajis02-348	81	46	ones	one	NOUN
ajis02-348	81	47	.	.	PUNCT
ajis02-348	82	1	on	on	ADP
ajis02-348	82	2	the	the	DET
ajis02-348	82	3	contrary	contrary	NOUN
ajis02-348	82	4	,	,	PUNCT
ajis02-348	82	5	the	the	DET
ajis02-348	82	6	weight	weight	NOUN
ajis02-348	82	7	of	of	ADP
ajis02-348	82	8	project	project	NOUN
ajis02-348	82	9	size	size	NOUN
ajis02-348	82	10	included	include	VERB
ajis02-348	82	11	in	in	ADP
ajis02-348	82	12	relative	relative	ADJ
ajis02-348	82	13	error	error	NOUN
ajis02-348	82	14	disappears	disappear	VERB
ajis02-348	82	15	from	from	ADP
ajis02-348	82	16	the	the	DET
ajis02-348	82	17	formula	formula	NOUN
ajis02-348	82	18	.	.	PUNCT
ajis02-348	83	1	so	so	ADV
ajis02-348	83	2	it	it	PRON
ajis02-348	83	3	falls	fall	VERB
ajis02-348	83	4	into	into	ADP
ajis02-348	83	5	the	the	DET
ajis02-348	83	6	category	category	NOUN
ajis02-348	83	7	of	of	ADP
ajis02-348	83	8	difference	difference	NOUN
ajis02-348	83	9	measure	measure	NOUN
ajis02-348	83	10	.	.	PUNCT
ajis02-348	84	1	occasionally	occasionally	ADV
ajis02-348	84	2	,	,	PUNCT
ajis02-348	84	3	a	a	DET
ajis02-348	84	4	combination	combination	NOUN
ajis02-348	84	5	of	of	ADP
ajis02-348	84	6	difference	difference	NOUN
ajis02-348	84	7	measure	measure	NOUN
ajis02-348	84	8	and	and	CCONJ
ajis02-348	84	9	ratio	ratio	NOUN
ajis02-348	84	10	measure	measure	NOUN
ajis02-348	84	11	is	be	AUX
ajis02-348	84	12	used	use	VERB
ajis02-348	84	13	for	for	ADP
ajis02-348	84	14	accuracy	accuracy	NOUN
ajis02-348	84	15	.	.	PUNCT
ajis02-348	85	1	j0rgensen	j0rgensen	PROPN
ajis02-348	85	2	(	(	PUNCT
ajis02-348	85	3	1994	1994	NUM
ajis02-348	85	4	)	)	PUNCT
ajis02-348	85	5	introduced	introduce	VERB
ajis02-348	85	6	pred2(0.25,0.5	pred2(0.25,0.5	NOUN
ajis02-348	85	7	)	)	PUNCT
ajis02-348	85	8	,	,	PUNCT
ajis02-348	85	9	the	the	DET
ajis02-348	85	10	percentage	percentage	NOUN
ajis02-348	85	11	of	of	ADP
ajis02-348	85	12	projects	project	NOUN
ajis02-348	85	13	with	with	ADP
ajis02-348	85	14	mre	mre	PROPN
ajis02-348	85	15	<	<	X
ajis02-348	85	16	0.25	0.25	NUM
ajis02-348	85	17	or	or	CCONJ
ajis02-348	85	18	estimation	estimation	NOUN
ajis02-348	85	19	error	error	NOUN
ajis02-348	85	20	<	<	X
ajis02-348	85	21	0.5	0.5	NUM
ajis02-348	85	22	days	day	NOUN
ajis02-348	85	23	.	.	PUNCT
ajis02-348	86	1	pred2(0.25,0.5	pred2(0.25,0.5	NOUN
ajis02-348	86	2	)	)	PUNCT
ajis02-348	86	3	is	be	AUX
ajis02-348	86	4	actually	actually	ADV
ajis02-348	86	5	pred(0.25	pred(0.25	ADJ
ajis02-348	86	6	)	)	PUNCT
ajis02-348	86	7	,	,	PUNCT
ajis02-348	86	8	because	because	SCONJ
ajis02-348	86	9	"	"	PUNCT
ajis02-348	86	10	estimation	estimation	NOUN
ajis02-348	86	11	error	error	NOUN
ajis02-348	86	12	<	<	X
ajis02-348	86	13	0.5	0.5	NUM
ajis02-348	86	14	days	day	NOUN
ajis02-348	86	15	"	"	PUNCT
ajis02-348	86	16	only	only	ADV
ajis02-348	86	17	contributes	contribute	VERB
ajis02-348	86	18	to	to	ADP
ajis02-348	86	19	the	the	DET
ajis02-348	86	20	percentage	percentage	NOUN
ajis02-348	86	21	when	when	SCONJ
ajis02-348	86	22	the	the	DET
ajis02-348	86	23	estimation	estimation	NOUN
ajis02-348	86	24	is	be	AUX
ajis02-348	86	25	performed	perform	VERB
ajis02-348	86	26	for	for	ADP
ajis02-348	86	27	a	a	DET
ajis02-348	86	28	project	project	NOUN
ajis02-348	86	29	of	of	ADP
ajis02-348	86	30	less	less	ADJ
ajis02-348	86	31	than	than	ADP
ajis02-348	86	32	two	two	NUM
ajis02-348	86	33	days	day	NOUN
ajis02-348	86	34	.	.	PUNCT
ajis02-348	87	1	discussion	discussion	NOUN
ajis02-348	87	2	on	on	ADP
ajis02-348	87	3	measures	measure	NOUN
ajis02-348	87	4	of	of	ADP
ajis02-348	87	5	accuracy	accuracy	NOUN
ajis02-348	87	6	kitchenham	kitchenham	INTJ
ajis02-348	87	7	et	et	PROPN
ajis02-348	87	8	al	al	PROPN
ajis02-348	87	9	(	(	PUNCT
ajis02-348	87	10	1990	1990	NUM
ajis02-348	87	11	)	)	PUNCT
ajis02-348	87	12	indicated	indicate	VERB
ajis02-348	87	13	that	that	SCONJ
ajis02-348	87	14	we	we	PRON
ajis02-348	87	15	should	should	AUX
ajis02-348	87	16	distinguish	distinguish	VERB
ajis02-348	87	17	between	between	ADP
ajis02-348	87	18	̂	̂	VERB
ajis02-348	87	19	z/	z/	NUM
ajis02-348	87	20	accuracy	accuracy	NOUN
ajis02-348	87	21	and	and	CCONJ
ajis02-348	87	22	prediction	prediction	NOUN
ajis02-348	87	23	accuracy	accuracy	NOUN
ajis02-348	87	24	,	,	PUNCT
ajis02-348	87	25	i.e.	i.e.	X
ajis02-348	87	26	how	how	SCONJ
ajis02-348	87	27	well	well	ADV
ajis02-348	87	28	a	a	DET
ajis02-348	87	29	model	model	NOUN
ajis02-348	87	30	fits	fit	VERB
ajis02-348	87	31	the	the	DET
ajis02-348	87	32	data	datum	NOUN
ajis02-348	87	33	from	from	ADP
ajis02-348	87	34	which	which	PRON
ajis02-348	87	35	it	it	PRON
ajis02-348	87	36	was	be	AUX
ajis02-348	87	37	generated	generate	VERB
ajis02-348	87	38	and	and	CCONJ
ajis02-348	87	39	how	how	SCONJ
ajis02-348	87	40	good	good	ADJ
ajis02-348	87	41	a	a	DET
ajis02-348	87	42	prediction	prediction	NOUN
ajis02-348	87	43	from	from	ADP
ajis02-348	87	44	the	the	DET
ajis02-348	87	45	model	model	NOUN
ajis02-348	87	46	will	will	AUX
ajis02-348	87	47	be	be	AUX
ajis02-348	87	48	.	.	PUNCT
ajis02-348	88	1	to	to	PART
ajis02-348	88	2	assess	assess	VERB
ajis02-348	88	3	fit	fit	ADJ
ajis02-348	88	4	accuracy	accuracy	NOUN
ajis02-348	88	5	,	,	PUNCT
ajis02-348	88	6	r	r	NOUN
ajis02-348	88	7	is	be	AUX
ajis02-348	88	8	a	a	DET
ajis02-348	88	9	good	good	ADJ
ajis02-348	88	10	measure	measure	NOUN
ajis02-348	88	11	.	.	PUNCT
ajis02-348	89	1	it	it	PRON
ajis02-348	89	2	can	can	AUX
ajis02-348	89	3	be	be	AUX
ajis02-348	89	4	seen	see	VERB
ajis02-348	89	5	from	from	ADP
ajis02-348	89	6	the	the	DET
ajis02-348	89	7	definition	definition	NOUN
ajis02-348	89	8	of	of	ADP
ajis02-348	89	9	r	r	NOUN
ajis02-348	89	10	that	that	SCONJ
ajis02-348	89	11	it	it	PRON
ajis02-348	89	12	is	be	AUX
ajis02-348	89	13	related	relate	VERB
ajis02-348	89	14	to	to	ADP
ajis02-348	89	15	the	the	DET
ajis02-348	89	16	variance	variance	NOUN
ajis02-348	89	17	of	of	ADP
ajis02-348	89	18	actual	actual	ADJ
ajis02-348	89	19	observed	observed	ADJ
ajis02-348	89	20	values	value	NOUN
ajis02-348	89	21	.	.	PUNCT
ajis02-348	90	1	the	the	PRON
ajis02-348	90	2	larger	large	ADJ
ajis02-348	90	3	this	this	DET
ajis02-348	90	4	variance	variance	NOUN
ajis02-348	90	5	is	be	AUX
ajis02-348	90	6	,	,	PUNCT
ajis02-348	90	7	the	the	PRON
ajis02-348	90	8	easier	easy	ADJ
ajis02-348	90	9	it	it	PRON
ajis02-348	90	10	is	be	AUX
ajis02-348	90	11	to	to	PART
ajis02-348	90	12	obtain	obtain	VERB
ajis02-348	90	13	a	a	DET
ajis02-348	90	14	large	large	ADJ
ajis02-348	90	15	r	r	NOUN
ajis02-348	90	16	.	.	PUNCT
ajis02-348	91	1	r	r	NOUN
ajis02-348	91	2	measures	measure	VERB
ajis02-348	91	3	the	the	DET
ajis02-348	91	4	proportion	proportion	NOUN
ajis02-348	91	5	of	of	ADP
ajis02-348	91	6	the	the	DET
ajis02-348	91	7	variation	variation	NOUN
ajis02-348	91	8	in	in	ADP
ajis02-348	91	9	the	the	DET
ajis02-348	91	10	dependent	dependent	ADJ
ajis02-348	91	11	variable	variable	NOUN
ajis02-348	91	12	that	that	PRON
ajis02-348	91	13	is	be	AUX
ajis02-348	91	14	explained	explain	VERB
ajis02-348	91	15	by	by	ADP
ajis02-348	91	16	the	the	DET
ajis02-348	91	17	regression	regression	NOUN
ajis02-348	91	18	equation	equation	NOUN
ajis02-348	91	19	(	(	PUNCT
ajis02-348	91	20	kenkel	kenkel	NOUN
ajis02-348	91	21	,	,	PUNCT
ajis02-348	91	22	1989	1989	NUM
ajis02-348	91	23	)	)	PUNCT
ajis02-348	91	24	.	.	PUNCT
ajis02-348	92	1	the	the	PRON
ajis02-348	92	2	higher	high	ADJ
ajis02-348	92	3	the	the	DET
ajis02-348	92	4	value	value	NOUN
ajis02-348	92	5	of	of	ADP
ajis02-348	92	6	r	r	NOUN
ajis02-348	92	7	,	,	PUNCT
ajis02-348	92	8	the	the	PRON
ajis02-348	92	9	greater	great	ADJ
ajis02-348	92	10	the	the	DET
ajis02-348	92	11	explanatory	explanatory	ADJ
ajis02-348	92	12	power	power	NOUN
ajis02-348	92	13	of	of	ADP
ajis02-348	92	14	the	the	DET
ajis02-348	92	15	regression	regression	NOUN
ajis02-348	92	16	equation	equation	NOUN
ajis02-348	92	17	.	.	PUNCT
ajis02-348	93	1	when	when	SCONJ
ajis02-348	93	2	r	r	NOUN
ajis02-348	93	3	is	be	AUX
ajis02-348	93	4	close	close	ADJ
ajis02-348	93	5	to	to	ADP
ajis02-348	93	6	0	0	NUM
ajis02-348	93	7	,	,	PUNCT
ajis02-348	93	8	either	either	CCONJ
ajis02-348	93	9	the	the	DET
ajis02-348	93	10	functional	functional	ADJ
ajis02-348	93	11	form	form	NOUN
ajis02-348	93	12	of	of	ADP
ajis02-348	93	13	the	the	DET
ajis02-348	93	14	model	model	NOUN
ajis02-348	93	15	does	do	AUX
ajis02-348	93	16	not	not	PART
ajis02-348	93	17	fit	fit	VERB
ajis02-348	93	18	the	the	DET
ajis02-348	93	19	data	datum	NOUN
ajis02-348	93	20	set	set	VERB
ajis02-348	93	21	from	from	ADP
ajis02-348	93	22	which	which	PRON
ajis02-348	93	23	it	it	PRON
ajis02-348	93	24	was	be	AUX
ajis02-348	93	25	generated	generate	VERB
ajis02-348	93	26	or	or	CCONJ
ajis02-348	93	27	more	more	ADV
ajis02-348	93	28	independent	independent	ADJ
ajis02-348	93	29	variables	variable	NOUN
ajis02-348	93	30	are	be	AUX
ajis02-348	93	31	needed	need	VERB
ajis02-348	93	32	to	to	PART
ajis02-348	93	33	further	far	ADV
ajis02-348	93	34	explain	explain	VERB
ajis02-348	93	35	the	the	DET
ajis02-348	93	36	variation	variation	NOUN
ajis02-348	93	37	in	in	ADP
ajis02-348	93	38	the	the	DET
ajis02-348	93	39	dependent	dependent	ADJ
ajis02-348	93	40	variable	variable	NOUN
ajis02-348	93	41	.	.	PUNCT
ajis02-348	94	1	to	to	PART
ajis02-348	94	2	assess	assess	VERB
ajis02-348	94	3	prediction	prediction	NOUN
ajis02-348	94	4	accuracy	accuracy	NOUN
ajis02-348	94	5	,	,	PUNCT
ajis02-348	94	6	difference	difference	NOUN
ajis02-348	94	7	measures	measure	NOUN
ajis02-348	94	8	are	be	AUX
ajis02-348	94	9	not	not	PART
ajis02-348	94	10	suitable	suitable	ADJ
ajis02-348	94	11	.	.	PUNCT
ajis02-348	95	1	with	with	ADP
ajis02-348	95	2	respect	respect	NOUN
ajis02-348	95	3	to	to	ADP
ajis02-348	95	4	software	software	NOUN
ajis02-348	95	5	cost	cost	NOUN
ajis02-348	95	6	estimation	estimation	NOUN
ajis02-348	95	7	,	,	PUNCT
ajis02-348	95	8	the	the	DET
ajis02-348	95	9	prediction	prediction	NOUN
ajis02-348	95	10	error	error	NOUN
ajis02-348	95	11	increases	increase	VERB
ajis02-348	95	12	with	with	ADP
ajis02-348	95	13	the	the	DET
ajis02-348	95	14	magnitude	magnitude	NOUN
ajis02-348	95	15	of	of	ADP
ajis02-348	95	16	the	the	DET
ajis02-348	95	17	observed	observed	ADJ
ajis02-348	95	18	value	value	NOUN
ajis02-348	95	19	.	.	PUNCT
ajis02-348	96	1	the	the	PRON
ajis02-348	96	2	larger	large	ADJ
ajis02-348	96	3	the	the	DET
ajis02-348	96	4	project	project	NOUN
ajis02-348	96	5	is	be	AUX
ajis02-348	96	6	,	,	PUNCT
ajis02-348	96	7	the	the	PRON
ajis02-348	96	8	harder	hard	ADJ
ajis02-348	96	9	it	it	PRON
ajis02-348	96	10	is	be	AUX
ajis02-348	96	11	to	to	PART
ajis02-348	96	12	estimate	estimate	VERB
ajis02-348	96	13	the	the	DET
ajis02-348	96	14	effort	effort	NOUN
ajis02-348	96	15	.	.	PUNCT
ajis02-348	97	1	difference	difference	NOUN
ajis02-348	97	2	measures	measure	NOUN
ajis02-348	97	3	are	be	AUX
ajis02-348	97	4	not	not	PART
ajis02-348	97	5	adequate	adequate	ADJ
ajis02-348	97	6	if	if	SCONJ
ajis02-348	97	7	they	they	PRON
ajis02-348	97	8	are	be	AUX
ajis02-348	97	9	used	use	VERB
ajis02-348	97	10	for	for	ADP
ajis02-348	97	11	both	both	CCONJ
ajis02-348	97	12	large	large	ADJ
ajis02-348	97	13	projects	project	NOUN
ajis02-348	97	14	and	and	CCONJ
ajis02-348	97	15	small	small	ADJ
ajis02-348	97	16	ones	one	NOUN
ajis02-348	97	17	,	,	PUNCT
ajis02-348	97	18	because	because	SCONJ
ajis02-348	97	19	they	they	PRON
ajis02-348	97	20	do	do	AUX
ajis02-348	97	21	not	not	PART
ajis02-348	97	22	take	take	VERB
ajis02-348	97	23	into	into	ADP
ajis02-348	97	24	consideration	consideration	NOUN
ajis02-348	97	25	the	the	DET
ajis02-348	97	26	size	size	NOUN
ajis02-348	97	27	of	of	ADP
ajis02-348	97	28	projects	project	NOUN
ajis02-348	97	29	.	.	PUNCT
ajis02-348	98	1	therefore	therefore	ADV
ajis02-348	98	2	difference	difference	NOUN
ajis02-348	98	3	measures	measure	NOUN
ajis02-348	98	4	should	should	AUX
ajis02-348	98	5	not	not	PART
ajis02-348	98	6	be	be	AUX
ajis02-348	98	7	used	use	VERB
ajis02-348	98	8	to	to	PART
ajis02-348	98	9	assess	assess	VERB
ajis02-348	98	10	the	the	DET
ajis02-348	98	11	prediction	prediction	NOUN
ajis02-348	98	12	accuracy	accuracy	NOUN
ajis02-348	98	13	for	for	SCONJ
ajis02-348	98	14	they	they	PRON
ajis02-348	98	15	penalise	penalise	VERB
ajis02-348	98	16	the	the	DET
ajis02-348	98	17	prediction	prediction	NOUN
ajis02-348	98	18	for	for	ADP
ajis02-348	98	19	large	large	ADJ
ajis02-348	98	20	projects	project	NOUN
ajis02-348	98	21	.	.	PUNCT
ajis02-348	99	1	on	on	ADP
ajis02-348	99	2	the	the	DET
ajis02-348	99	3	other	other	ADJ
ajis02-348	99	4	hand	hand	NOUN
ajis02-348	99	5	,	,	PUNCT
ajis02-348	99	6	relative	relative	ADJ
ajis02-348	99	7	error	error	NOUN
ajis02-348	99	8	is	be	AUX
ajis02-348	99	9	an	an	DET
ajis02-348	99	10	average	average	NOUN
ajis02-348	99	11	of	of	ADP
ajis02-348	99	12	prediction	prediction	NOUN
ajis02-348	99	13	error	error	NOUN
ajis02-348	99	14	in	in	ADP
ajis02-348	99	15	every	every	DET
ajis02-348	99	16	unit	unit	NOUN
ajis02-348	99	17	of	of	ADP
ajis02-348	99	18	effort	effort	NOUN
ajis02-348	99	19	,	,	PUNCT
ajis02-348	99	20	which	which	PRON
ajis02-348	99	21	reflects	reflect	VERB
ajis02-348	99	22	"	"	PUNCT
ajis02-348	99	23	error	error	NOUN
ajis02-348	99	24	rate	rate	NOUN
ajis02-348	99	25	"	"	PUNCT
ajis02-348	99	26	.	.	PUNCT
ajis02-348	100	1	it	it	PRON
ajis02-348	100	2	takes	take	VERB
ajis02-348	100	3	into	into	ADP
ajis02-348	100	4	account	account	NOUN
ajis02-348	100	5	the	the	DET
ajis02-348	100	6	project	project	NOUN
ajis02-348	100	7	size	size	NOUN
ajis02-348	100	8	and	and	CCONJ
ajis02-348	100	9	allows	allow	VERB
ajis02-348	100	10	larger	large	ADJ
ajis02-348	100	11	absolute	absolute	ADJ
ajis02-348	100	12	error	error	NOUN
ajis02-348	100	13	for	for	ADP
ajis02-348	100	14	larger	large	ADJ
ajis02-348	100	15	projects	project	NOUN
ajis02-348	100	16	.	.	PUNCT
ajis02-348	101	1	therefore	therefore	ADV
ajis02-348	101	2	,	,	PUNCT
ajis02-348	101	3	ratio	ratio	NOUN
ajis02-348	101	4	measure	measure	NOUN
ajis02-348	101	5	is	be	AUX
ajis02-348	101	6	more	more	ADV
ajis02-348	101	7	suitable	suitable	ADJ
ajis02-348	101	8	for	for	ADP
ajis02-348	101	9	the	the	DET
ajis02-348	101	10	assessment	assessment	NOUN
ajis02-348	101	11	of	of	ADP
ajis02-348	101	12	the	the	DET
ajis02-348	101	13	accuracy	accuracy	NOUN
ajis02-348	101	14	in	in	ADP
ajis02-348	101	15	software	software	NOUN
ajis02-348	101	16	cost	cost	NOUN
ajis02-348	101	17	estimation	estimation	NOUN
ajis02-348	101	18	.	.	PUNCT
ajis02-348	102	1	among	among	ADP
ajis02-348	102	2	the	the	DET
ajis02-348	102	3	ratio	ratio	NOUN
ajis02-348	102	4	measures	measure	NOUN
ajis02-348	102	5	,	,	PUNCT
ajis02-348	102	6	all	all	PRON
ajis02-348	102	7	mean	mean	NOUN
ajis02-348	102	8	measures	measure	NOUN
ajis02-348	102	9	are	be	AUX
ajis02-348	102	10	influenced	influence	VERB
ajis02-348	102	11	by	by	ADP
ajis02-348	102	12	very	very	ADV
ajis02-348	102	13	large	large	ADJ
ajis02-348	102	14	extreme	extreme	ADJ
ajis02-348	102	15	mre	mre	PROPN
ajis02-348	102	16	values	value	NOUN
ajis02-348	102	17	,	,	PUNCT
ajis02-348	102	18	which	which	PRON
ajis02-348	102	19	usually	usually	ADV
ajis02-348	102	20	occur	occur	VERB
ajis02-348	102	21	in	in	ADP
ajis02-348	102	22	software	software	NOUN
ajis02-348	102	23	estimation	estimation	NOUN
ajis02-348	102	24	.	.	PUNCT
ajis02-348	103	1	in	in	ADP
ajis02-348	103	2	the	the	DET
ajis02-348	103	3	context	context	NOUN
ajis02-348	103	4	of	of	ADP
ajis02-348	103	5	accuracy	accuracy	NOUN
ajis02-348	103	6	evaluation	evaluation	NOUN
ajis02-348	103	7	for	for	ADP
ajis02-348	103	8	software	software	NOUN
ajis02-348	103	9	cost	cost	NOUN
ajis02-348	103	10	estimation	estimation	NOUN
ajis02-348	103	11	,	,	PUNCT
ajis02-348	103	12	they	they	PRON
ajis02-348	103	13	can	can	AUX
ajis02-348	103	14	not	not	PART
ajis02-348	103	15	play	play	VERB
ajis02-348	103	16	the	the	DET
ajis02-348	103	17	role	role	NOUN
ajis02-348	103	18	of	of	ADP
ajis02-348	103	19	the	the	DET
ajis02-348	103	20	measure	measure	NOUN
ajis02-348	103	21	for	for	ADP
ajis02-348	103	22	central	central	ADJ
ajis02-348	103	23	tendency	tendency	NOUN
ajis02-348	103	24	.	.	PUNCT
ajis02-348	104	1	some	some	PRON
ajis02-348	104	2	of	of	ADP
ajis02-348	104	3	them	they	PRON
ajis02-348	104	4	favour	favour	VERB
ajis02-348	104	5	underestimation	underestimation	NOUN
ajis02-348	104	6	.	.	PUNCT
ajis02-348	105	1	for	for	ADP
ajis02-348	105	2	the	the	DET
ajis02-348	105	3	single	single	ADJ
ajis02-348	105	4	value	value	NOUN
ajis02-348	105	5	measures	measure	NOUN
ajis02-348	105	6	,	,	PUNCT
ajis02-348	105	7	although	although	SCONJ
ajis02-348	105	8	they	they	PRON
ajis02-348	105	9	eliminate	eliminate	VERB
ajis02-348	105	10	the	the	DET
ajis02-348	105	11	influence	influence	NOUN
ajis02-348	105	12	of	of	ADP
ajis02-348	105	13	outliers	outlier	NOUN
ajis02-348	105	14	,	,	PUNCT
ajis02-348	105	15	they	they	PRON
ajis02-348	105	16	can	can	AUX
ajis02-348	105	17	hardly	hardly	ADV
ajis02-348	105	18	provide	provide	VERB
ajis02-348	105	19	an	an	DET
ajis02-348	105	20	overall	overall	ADJ
ajis02-348	105	21	accuracy	accuracy	NOUN
ajis02-348	105	22	.	.	PUNCT
ajis02-348	106	1	for	for	ADP
ajis02-348	106	2	example	example	NOUN
ajis02-348	106	3	,	,	PUNCT
ajis02-348	106	4	pred(0.25	pred(0.25	NOUN
ajis02-348	106	5	)	)	PUNCT
ajis02-348	106	6	takes	take	VERB
ajis02-348	106	7	into	into	ADP
ajis02-348	106	8	account	account	NOUN
ajis02-348	106	9	less	less	ADJ
ajis02-348	106	10	than	than	ADP
ajis02-348	106	11	75	75	NUM
ajis02-348	106	12	%	%	NOUN
ajis02-348	106	13	of	of	ADP
ajis02-348	106	14	all	all	DET
ajis02-348	106	15	the	the	DET
ajis02-348	106	16	wire	wire	NOUN
ajis02-348	106	17	's	's	PART
ajis02-348	106	18	,	,	PUNCT
ajis02-348	106	19	because	because	SCONJ
ajis02-348	106	20	conte	conte	PROPN
ajis02-348	106	21	's	's	PART
ajis02-348	106	22	acceptable	acceptable	ADJ
ajis02-348	106	23	level	level	NOUN
ajis02-348	106	24	can	can	AUX
ajis02-348	106	25	seldom	seldom	ADV
ajis02-348	106	26	be	be	AUX
ajis02-348	106	27	achieved	achieve	VERB
ajis02-348	106	28	.	.	PUNCT
ajis02-348	107	1	although	although	SCONJ
ajis02-348	107	2	q3	q3	PROPN
ajis02-348	107	3	indicates	indicate	VERB
ajis02-348	107	4	that	that	SCONJ
ajis02-348	107	5	75	75	NUM
ajis02-348	107	6	%	%	NOUN
ajis02-348	107	7	of	of	ADP
ajis02-348	107	8	the	the	DET
ajis02-348	107	9	projects	project	NOUN
ajis02-348	107	10	are	be	AUX
ajis02-348	107	11	estimated	estimate	VERB
ajis02-348	107	12	with	with	ADP
ajis02-348	107	13	mre	mre	PROPN
ajis02-348	107	14	less	less	ADJ
ajis02-348	107	15	than	than	ADP
ajis02-348	107	16	or	or	CCONJ
ajis02-348	107	17	equal	equal	ADJ
ajis02-348	107	18	to	to	ADP
ajis02-348	107	19	<	<	X
ajis02-348	107	20	23	23	NUM
ajis02-348	107	21	,	,	PUNCT
ajis02-348	107	22	it	it	PRON
ajis02-348	107	23	does	do	AUX
ajis02-348	107	24	not	not	PART
ajis02-348	107	25	provide	provide	VERB
ajis02-348	107	26	any	any	DET
ajis02-348	107	27	detailed	detailed	ADJ
ajis02-348	107	28	information	information	NOUN
ajis02-348	107	29	on	on	ADP
ajis02-348	107	30	these	these	DET
ajis02-348	107	31	mre	mre	PROPN
ajis02-348	107	32	values	value	NOUN
ajis02-348	107	33	.	.	PUNCT
ajis02-348	108	1	for	for	ADP
ajis02-348	108	2	the	the	DET
ajis02-348	108	3	purpose	purpose	NOUN
ajis02-348	108	4	of	of	ADP
ajis02-348	108	5	comparison	comparison	NOUN
ajis02-348	108	6	among	among	ADP
ajis02-348	108	7	different	different	ADJ
ajis02-348	108	8	models	model	NOUN
ajis02-348	108	9	,	,	PUNCT
ajis02-348	108	10	the	the	DET
ajis02-348	108	11	single	single	ADJ
ajis02-348	108	12	value	value	NOUN
ajis02-348	108	13	measures	measure	NOUN
ajis02-348	108	14	are	be	AUX
ajis02-348	108	15	rough	rough	ADJ
ajis02-348	108	16	and	and	CCONJ
ajis02-348	108	17	stochastic	stochastic	ADJ
ajis02-348	108	18	.	.	PUNCT
ajis02-348	109	1	moreover	moreover	ADV
ajis02-348	109	2	,	,	PUNCT
ajis02-348	109	3	these	these	DET
ajis02-348	109	4	single	single	ADJ
ajis02-348	109	5	value	value	NOUN
ajis02-348	109	6	measures	measure	NOUN
ajis02-348	109	7	and	and	CCONJ
ajis02-348	109	8	mean	mean	ADJ
ajis02-348	109	9	measures	measure	NOUN
ajis02-348	109	10	are	be	AUX
ajis02-348	109	11	not	not	PART
ajis02-348	109	12	always	always	ADV
ajis02-348	109	13	consistent	consistent	ADJ
ajis02-348	109	14	when	when	SCONJ
ajis02-348	109	15	they	they	PRON
ajis02-348	109	16	are	be	AUX
ajis02-348	109	17	used	use	VERB
ajis02-348	109	18	to	to	PART
ajis02-348	109	19	compare	compare	VERB
ajis02-348	109	20	different	different	ADJ
ajis02-348	109	21	models	model	NOUN
ajis02-348	109	22	.	.	PUNCT
ajis02-348	110	1	ajis	ajis	ADV
ajis02-348	110	2	vol	vol	VERB
ajis02-348	110	3	5	5	NUM
ajis02-348	110	4	no	no	DET
ajis02-348	110	5	1	1	NUM
ajis02-348	110	6	september	september	NOUN
ajis02-348	110	7	1997	1997	NUM
ajis02-348	110	8	34	34	NUM
ajis02-348	110	9	the	the	DET
ajis02-348	110	10	above	above	ADJ
ajis02-348	110	11	discussion	discussion	NOUN
ajis02-348	110	12	may	may	AUX
ajis02-348	110	13	be	be	AUX
ajis02-348	110	14	summarized	summarize	VERB
ajis02-348	110	15	in	in	ADP
ajis02-348	110	16	three	three	NUM
ajis02-348	110	17	points	point	NOUN
ajis02-348	110	18	:	:	PUNCT
ajis02-348	110	19	•	•	SCONJ
ajis02-348	110	20	the	the	DET
ajis02-348	110	21	difference	difference	NOUN
ajis02-348	110	22	measure	measure	NOUN
ajis02-348	110	23	is	be	AUX
ajis02-348	110	24	not	not	PART
ajis02-348	110	25	suitable	suitable	ADJ
ajis02-348	110	26	for	for	ADP
ajis02-348	110	27	the	the	DET
ajis02-348	110	28	evaluation	evaluation	NOUN
ajis02-348	110	29	of	of	ADP
ajis02-348	110	30	software	software	NOUN
ajis02-348	110	31	cost	cost	NOUN
ajis02-348	110	32	models	model	NOUN
ajis02-348	110	33	.	.	PUNCT
ajis02-348	111	1	the	the	DET
ajis02-348	111	2	ratio	ratio	NOUN
ajis02-348	111	3	measure	measure	NOUN
ajis02-348	111	4	is	be	AUX
ajis02-348	111	5	preferable	preferable	ADJ
ajis02-348	111	6	.	.	PUNCT
ajis02-348	112	1	•	•	NOUN
ajis02-348	112	2	the	the	DET
ajis02-348	112	3	mean	mean	NOUN
ajis02-348	112	4	measures	measure	NOUN
ajis02-348	112	5	take	take	VERB
ajis02-348	112	6	into	into	ADP
ajis02-348	112	7	account	account	NOUN
ajis02-348	112	8	every	every	DET
ajis02-348	112	9	single	single	ADJ
ajis02-348	112	10	relative	relative	ADJ
ajis02-348	112	11	error	error	NOUN
ajis02-348	112	12	,	,	PUNCT
ajis02-348	112	13	but	but	CCONJ
ajis02-348	112	14	they	they	PRON
ajis02-348	112	15	are	be	AUX
ajis02-348	112	16	significantly	significantly	ADV
ajis02-348	112	17	influenced	influence	VERB
ajis02-348	112	18	by	by	ADP
ajis02-348	112	19	outliers	outlier	NOUN
ajis02-348	112	20	,	,	PUNCT
ajis02-348	112	21	which	which	PRON
ajis02-348	112	22	frequently	frequently	ADV
ajis02-348	112	23	occur	occur	VERB
ajis02-348	112	24	in	in	ADP
ajis02-348	112	25	software	software	NOUN
ajis02-348	112	26	cost	cost	NOUN
ajis02-348	112	27	estimation	estimation	NOUN
ajis02-348	112	28	.	.	PUNCT
ajis02-348	113	1	some	some	PRON
ajis02-348	113	2	of	of	ADP
ajis02-348	113	3	them	they	PRON
ajis02-348	113	4	favour	favour	VERB
ajis02-348	113	5	underestimation	underestimation	NOUN
ajis02-348	113	6	.	.	PUNCT
ajis02-348	114	1	•	•	NUM
ajis02-348	114	2	single	single	ADJ
ajis02-348	114	3	value	value	NOUN
ajis02-348	114	4	measures	measure	NOUN
ajis02-348	114	5	are	be	AUX
ajis02-348	114	6	not	not	PART
ajis02-348	114	7	influenced	influence	VERB
ajis02-348	114	8	by	by	ADP
ajis02-348	114	9	outliers	outlier	NOUN
ajis02-348	114	10	,	,	PUNCT
ajis02-348	114	11	but	but	CCONJ
ajis02-348	114	12	they	they	PRON
ajis02-348	114	13	are	be	AUX
ajis02-348	114	14	stochastic	stochastic	ADJ
ajis02-348	114	15	and	and	CCONJ
ajis02-348	114	16	lose	lose	VERB
ajis02-348	114	17	too	too	ADV
ajis02-348	114	18	much	much	ADJ
ajis02-348	114	19	information	information	NOUN
ajis02-348	114	20	.	.	PUNCT
ajis02-348	115	1	therefore	therefore	ADV
ajis02-348	115	2	the	the	DET
ajis02-348	115	3	evaluation	evaluation	NOUN
ajis02-348	115	4	by	by	ADP
ajis02-348	115	5	single	single	ADJ
ajis02-348	115	6	value	value	NOUN
ajis02-348	115	7	measures	measure	NOUN
ajis02-348	115	8	is	be	AUX
ajis02-348	115	9	not	not	PART
ajis02-348	115	10	always	always	ADV
ajis02-348	115	11	reliable	reliable	ADJ
ajis02-348	115	12	when	when	SCONJ
ajis02-348	115	13	they	they	PRON
ajis02-348	115	14	are	be	AUX
ajis02-348	115	15	used	use	VERB
ajis02-348	115	16	to	to	PART
ajis02-348	115	17	compare	compare	VERB
ajis02-348	115	18	different	different	ADJ
ajis02-348	115	19	models	model	NOUN
ajis02-348	115	20	.	.	PUNCT
ajis02-348	116	1	because	because	SCONJ
ajis02-348	116	2	of	of	ADP
ajis02-348	116	3	the	the	DET
ajis02-348	116	4	above	above	ADJ
ajis02-348	116	5	drawbacks	drawback	NOUN
ajis02-348	116	6	of	of	ADP
ajis02-348	116	7	the	the	DET
ajis02-348	116	8	existing	exist	VERB
ajis02-348	116	9	accuracy	accuracy	NOUN
ajis02-348	116	10	measures	measure	NOUN
ajis02-348	116	11	,	,	PUNCT
ajis02-348	116	12	there	there	PRON
ajis02-348	116	13	is	be	VERB
ajis02-348	116	14	a	a	DET
ajis02-348	116	15	need	need	NOUN
ajis02-348	116	16	for	for	ADP
ajis02-348	116	17	a	a	DET
ajis02-348	116	18	better	well	ADJ
ajis02-348	116	19	measure	measure	NOUN
ajis02-348	116	20	for	for	ADP
ajis02-348	116	21	accuracy	accuracy	NOUN
ajis02-348	116	22	evaluation	evaluation	NOUN
ajis02-348	116	23	.	.	PUNCT
ajis02-348	117	1	a	a	DET
ajis02-348	117	2	new	new	ADJ
ajis02-348	117	3	measure	measure	NOUN
ajis02-348	117	4	of	of	ADP
ajis02-348	117	5	accuracy	accuracy	NOUN
ajis02-348	117	6	weighted	weight	VERB
ajis02-348	117	7	mean	mean	NOUN
ajis02-348	117	8	of	of	ADP
ajis02-348	117	9	quartiles	quartile	NOUN
ajis02-348	117	10	of	of	ADP
ajis02-348	117	11	mre	mre	PROPN
ajis02-348	117	12	's	's	PART
ajis02-348	117	13	(	(	PUNCT
ajis02-348	117	14	wmq	wmq	NOUN
ajis02-348	117	15	)	)	PUNCT
ajis02-348	117	16	to	to	PART
ajis02-348	117	17	resolve	resolve	VERB
ajis02-348	117	18	the	the	DET
ajis02-348	117	19	above	above	ADJ
ajis02-348	117	20	problems	problem	NOUN
ajis02-348	117	21	,	,	PUNCT
ajis02-348	117	22	it	it	PRON
ajis02-348	117	23	is	be	AUX
ajis02-348	117	24	proposed	propose	VERB
ajis02-348	117	25	to	to	PART
ajis02-348	117	26	use	use	VERB
ajis02-348	117	27	the	the	DET
ajis02-348	117	28	weighted	weight	VERB
ajis02-348	117	29	mean	mean	NOUN
ajis02-348	117	30	of	of	ADP
ajis02-348	117	31	quartiles	quartile	NOUN
ajis02-348	117	32	(	(	PUNCT
ajis02-348	117	33	wmq	wmq	NOUN
ajis02-348	117	34	)	)	PUNCT
ajis02-348	117	35	of	of	ADP
ajis02-348	117	36	mre	mre	PROPN
ajis02-348	117	37	to	to	PART
ajis02-348	117	38	measure	measure	VERB
ajis02-348	117	39	the	the	DET
ajis02-348	117	40	prediction	prediction	NOUN
ajis02-348	117	41	accuracy	accuracy	NOUN
ajis02-348	117	42	.	.	PUNCT
ajis02-348	118	1	the	the	DET
ajis02-348	118	2	third	third	ADJ
ajis02-348	118	3	quartile	quartile	NOUN
ajis02-348	118	4	(	(	PUNCT
ajis02-348	118	5	q3	q3	PROPN
ajis02-348	118	6	)	)	PUNCT
ajis02-348	118	7	is	be	AUX
ajis02-348	118	8	the	the	DET
ajis02-348	118	9	most	most	ADV
ajis02-348	118	10	important	important	ADJ
ajis02-348	118	11	as	as	ADP
ajis02-348	118	12	75	75	NUM
ajis02-348	118	13	%	%	NOUN
ajis02-348	118	14	of	of	ADP
ajis02-348	118	15	mre	mre	PROPN
ajis02-348	118	16	's	's	PART
ajis02-348	118	17	values	value	NOUN
ajis02-348	118	18	are	be	AUX
ajis02-348	118	19	less	less	ADJ
ajis02-348	118	20	than	than	ADP
ajis02-348	118	21	it	it	PRON
ajis02-348	118	22	.	.	PUNCT
ajis02-348	119	1	so	so	ADV
ajis02-348	119	2	it	it	PRON
ajis02-348	119	3	is	be	AUX
ajis02-348	119	4	weighted	weight	VERB
ajis02-348	119	5	with	with	ADP
ajis02-348	119	6	75	75	NUM
ajis02-348	119	7	.	.	PUNCT
ajis02-348	120	1	the	the	DET
ajis02-348	120	2	second	second	ADJ
ajis02-348	120	3	quartile	quartile	NOUN
ajis02-348	120	4	(	(	PUNCT
ajis02-348	120	5	q2	q2	NOUN
ajis02-348	120	6	)	)	PUNCT
ajis02-348	120	7	is	be	AUX
ajis02-348	120	8	weighted	weight	VERB
ajis02-348	120	9	with	with	ADP
ajis02-348	120	10	50	50	NUM
ajis02-348	120	11	,	,	PUNCT
ajis02-348	120	12	and	and	CCONJ
ajis02-348	120	13	the	the	DET
ajis02-348	120	14	first	first	ADJ
ajis02-348	120	15	quartile	quartile	NOUN
ajis02-348	120	16	(	(	PUNCT
ajis02-348	120	17	q	q	NOUN
ajis02-348	120	18	,	,	PUNCT
ajis02-348	120	19	)	)	PUNCT
ajis02-348	120	20	is	be	AUX
ajis02-348	120	21	weighted	weight	VERB
ajis02-348	120	22	with	with	ADP
ajis02-348	120	23	25	25	NUM
ajis02-348	120	24	.	.	PUNCT
ajis02-348	121	1	the	the	DET
ajis02-348	121	2	wmq	wmq	NOUN
ajis02-348	121	3	is	be	AUX
ajis02-348	121	4	defined	define	VERB
ajis02-348	121	5	as	as	ADP
ajis02-348	121	6	:	:	PUNCT
ajis02-348	121	7	wmq	wmq	NOUN
ajis02-348	121	8	150	150	NUM
ajis02-348	121	9	6	6	NUM
ajis02-348	121	10	there	there	PRON
ajis02-348	121	11	are	be	VERB
ajis02-348	121	12	two	two	NUM
ajis02-348	121	13	assumptions	assumption	NOUN
ajis02-348	121	14	underlying	underlie	VERB
ajis02-348	121	15	the	the	DET
ajis02-348	121	16	use	use	NOUN
ajis02-348	121	17	of	of	ADP
ajis02-348	121	18	wmq	wmq	NOUN
ajis02-348	121	19	:	:	PUNCT
ajis02-348	121	20	1	1	X
ajis02-348	121	21	)	)	PUNCT
ajis02-348	121	22	the	the	DET
ajis02-348	121	23	number	number	NOUN
ajis02-348	121	24	of	of	ADP
ajis02-348	121	25	outliers	outlier	NOUN
ajis02-348	121	26	is	be	AUX
ajis02-348	121	27	less	less	ADJ
ajis02-348	121	28	than	than	ADP
ajis02-348	121	29	25	25	NUM
ajis02-348	121	30	%	%	NOUN
ajis02-348	121	31	of	of	ADP
ajis02-348	121	32	all	all	DET
ajis02-348	121	33	the	the	DET
ajis02-348	121	34	mre	mre	PROPN
ajis02-348	121	35	values	value	NOUN
ajis02-348	121	36	.	.	PUNCT
ajis02-348	122	1	this	this	DET
ajis02-348	122	2	assumption	assumption	NOUN
ajis02-348	122	3	is	be	AUX
ajis02-348	122	4	generally	generally	ADV
ajis02-348	122	5	true	true	ADJ
ajis02-348	122	6	.	.	PUNCT
ajis02-348	123	1	2	2	X
ajis02-348	123	2	)	)	PUNCT
ajis02-348	123	3	if	if	SCONJ
ajis02-348	123	4	the	the	DET
ajis02-348	123	5	estimation	estimation	NOUN
ajis02-348	123	6	of	of	ADP
ajis02-348	123	7	75	75	NUM
ajis02-348	123	8	%	%	NOUN
ajis02-348	123	9	of	of	ADP
ajis02-348	123	10	the	the	DET
ajis02-348	123	11	projects	project	NOUN
ajis02-348	123	12	is	be	AUX
ajis02-348	123	13	acceptable	acceptable	ADJ
ajis02-348	123	14	,	,	PUNCT
ajis02-348	123	15	the	the	DET
ajis02-348	123	16	model	model	NOUN
ajis02-348	123	17	is	be	AUX
ajis02-348	123	18	desirable	desirable	ADJ
ajis02-348	123	19	.	.	PUNCT
ajis02-348	124	1	because	because	SCONJ
ajis02-348	124	2	of	of	ADP
ajis02-348	124	3	the	the	DET
ajis02-348	124	4	present	present	ADJ
ajis02-348	124	5	low	low	ADJ
ajis02-348	124	6	level	level	NOUN
ajis02-348	124	7	of	of	ADP
ajis02-348	124	8	poor	poor	ADJ
ajis02-348	124	9	estimation	estimation	NOUN
ajis02-348	124	10	accuracy	accuracy	NOUN
ajis02-348	124	11	,	,	PUNCT
ajis02-348	124	12	this	this	DET
ajis02-348	124	13	assumption	assumption	NOUN
ajis02-348	124	14	is	be	AUX
ajis02-348	124	15	reasonable	reasonable	ADJ
ajis02-348	124	16	.	.	PUNCT
ajis02-348	125	1	unlike	unlike	ADP
ajis02-348	125	2	the	the	DET
ajis02-348	125	3	mean	mean	ADJ
ajis02-348	125	4	measures	measure	NOUN
ajis02-348	125	5	,	,	PUNCT
ajis02-348	125	6	the	the	DET
ajis02-348	125	7	wmq	wmq	NOUN
ajis02-348	125	8	is	be	AUX
ajis02-348	125	9	not	not	PART
ajis02-348	125	10	influenced	influence	VERB
ajis02-348	125	11	by	by	ADP
ajis02-348	125	12	extreme	extreme	ADJ
ajis02-348	125	13	mre	mre	PROPN
ajis02-348	125	14	values	value	NOUN
ajis02-348	125	15	,	,	PUNCT
ajis02-348	125	16	and	and	CCONJ
ajis02-348	125	17	at	at	ADP
ajis02-348	125	18	the	the	DET
ajis02-348	125	19	same	same	ADJ
ajis02-348	125	20	time	time	NOUN
ajis02-348	125	21	it	it	PRON
ajis02-348	125	22	provides	provide	VERB
ajis02-348	125	23	more	more	ADJ
ajis02-348	125	24	general	general	ADJ
ajis02-348	125	25	information	information	NOUN
ajis02-348	125	26	about	about	ADP
ajis02-348	125	27	the	the	DET
ajis02-348	125	28	distribution	distribution	NOUN
ajis02-348	125	29	of	of	ADP
ajis02-348	125	30	mre	mre	NOUN
ajis02-348	125	31	than	than	ADP
ajis02-348	125	32	the	the	DET
ajis02-348	125	33	single	single	ADJ
ajis02-348	125	34	value	value	NOUN
ajis02-348	125	35	measures	measure	NOUN
ajis02-348	125	36	.	.	PUNCT
ajis02-348	126	1	it	it	PRON
ajis02-348	126	2	will	will	AUX
ajis02-348	126	3	be	be	AUX
ajis02-348	126	4	shown	show	VERB
ajis02-348	126	5	later	later	ADV
ajis02-348	126	6	that	that	SCONJ
ajis02-348	126	7	the	the	DET
ajis02-348	126	8	wmq	wmq	NOUN
ajis02-348	126	9	presents	present	VERB
ajis02-348	126	10	a	a	DET
ajis02-348	126	11	good	good	ADJ
ajis02-348	126	12	average	average	NOUN
ajis02-348	126	13	of	of	ADP
ajis02-348	126	14	mre	mre	NOUN
ajis02-348	126	15	in	in	ADP
ajis02-348	126	16	evaluating	evaluate	VERB
ajis02-348	126	17	accuracy	accuracy	NOUN
ajis02-348	126	18	of	of	ADP
ajis02-348	126	19	software	software	NOUN
ajis02-348	126	20	cost	cost	NOUN
ajis02-348	126	21	estimation	estimation	NOUN
ajis02-348	126	22	.	.	PUNCT
ajis02-348	127	1	the	the	DET
ajis02-348	127	2	wmq	wmq	NOUN
ajis02-348	127	3	is	be	AUX
ajis02-348	127	4	consistent	consistent	ADJ
ajis02-348	127	5	with	with	ADP
ajis02-348	127	6	the	the	DET
ajis02-348	127	7	mre	mre	PROPN
ajis02-348	127	8	provided	provide	VERB
ajis02-348	127	9	that	that	SCONJ
ajis02-348	127	10	the	the	DET
ajis02-348	127	11	estimation	estimation	NOUN
ajis02-348	127	12	is	be	AUX
ajis02-348	127	13	not	not	PART
ajis02-348	127	14	obviously	obviously	ADV
ajis02-348	127	15	biased	bias	VERB
ajis02-348	127	16	(	(	PUNCT
ajis02-348	127	17	tending	tend	VERB
ajis02-348	127	18	to	to	ADP
ajis02-348	127	19	under	under	ADP
ajis02-348	127	20	/	/	SYM
ajis02-348	127	21	overestimate	overestimate	NOUN
ajis02-348	127	22	)	)	PUNCT
ajis02-348	127	23	and	and	CCONJ
ajis02-348	127	24	there	there	PRON
ajis02-348	127	25	are	be	VERB
ajis02-348	127	26	no	no	DET
ajis02-348	127	27	outliers	outlier	NOUN
ajis02-348	127	28	(	(	PUNCT
ajis02-348	127	29	extreme	extreme	ADJ
ajis02-348	127	30	large	large	ADJ
ajis02-348	127	31	values	value	NOUN
ajis02-348	127	32	)	)	PUNCT
ajis02-348	127	33	.	.	PUNCT
ajis02-348	128	1	in	in	ADP
ajis02-348	128	2	a	a	DET
ajis02-348	128	3	sample	sample	NOUN
ajis02-348	128	4	of	of	ADP
ajis02-348	128	5	n	n	PRON
ajis02-348	128	6	projects	project	VERB
ajis02-348	128	7	an	an	DET
ajis02-348	128	8	estimation	estimation	NOUN
ajis02-348	128	9	model	model	NOUN
ajis02-348	128	10	results	result	NOUN
ajis02-348	128	11	in	in	ADP
ajis02-348	128	12	n	n	CCONJ
ajis02-348	128	13	relative	relative	ADJ
ajis02-348	128	14	errors	error	NOUN
ajis02-348	128	15	,	,	PUNCT
ajis02-348	128	16	the	the	DET
ajis02-348	128	17	magnitude	magnitude	NOUN
ajis02-348	128	18	of	of	ADP
ajis02-348	128	19	which	which	PRON
ajis02-348	128	20	is	be	AUX
ajis02-348	128	21	expressed	express	VERB
ajis02-348	128	22	as	as	ADP
ajis02-348	128	23	a	a	DET
ajis02-348	128	24	variable	variable	ADJ
ajis02-348	128	25	mre	mre	NOUN
ajis02-348	128	26	.	.	PUNCT
ajis02-348	129	1	these	these	DET
ajis02-348	129	2	mre	mre	PROPN
ajis02-348	129	3	's	's	PART
ajis02-348	129	4	ait	ait	NOUN
ajis02-348	129	5	arranged	arrange	VERB
ajis02-348	129	6	in	in	ADP
ajis02-348	129	7	ascending	ascend	VERB
ajis02-348	129	8	order	order	NOUN
ajis02-348	129	9	.	.	PUNCT
ajis02-348	130	1	the	the	DET
ajis02-348	130	2	mre	mre	PROPN
ajis02-348	130	3	can	can	AUX
ajis02-348	130	4	be	be	AUX
ajis02-348	130	5	expressed	express	VERB
ajis02-348	130	6	as	as	ADP
ajis02-348	130	7	a	a	DET
ajis02-348	130	8	function	function	NOUN
ajis02-348	130	9	of	of	ADP
ajis02-348	130	10	the	the	DET
ajis02-348	130	11	order	order	NOUN
ajis02-348	130	12	number	number	NOUN
ajis02-348	130	13	n	n	CCONJ
ajis02-348	130	14	,	,	PUNCT
ajis02-348	130	15	i.e.	i.e.	X
ajis02-348	130	16	mre	mre	PROPN
ajis02-348	130	17	=	=	NOUN
ajis02-348	130	18	f(n	f(n	PROPN
ajis02-348	130	19	)	)	PUNCT
ajis02-348	130	20	with	with	ADP
ajis02-348	130	21	this	this	DET
ajis02-348	130	22	equation	equation	NOUN
ajis02-348	130	23	a	a	DET
ajis02-348	130	24	smoothed	smoothed	ADJ
ajis02-348	130	25	curve	curve	NOUN
ajis02-348	130	26	(	(	PUNCT
ajis02-348	130	27	oq^q2q^q4	oq^q2q^q4	ADJ
ajis02-348	130	28	)	)	PUNCT
ajis02-348	130	29	can	can	AUX
ajis02-348	130	30	be	be	AUX
ajis02-348	130	31	drawn	draw	VERB
ajis02-348	130	32	.	.	PUNCT
ajis02-348	131	1	this	this	DET
ajis02-348	131	2	curve	curve	NOUN
ajis02-348	131	3	is	be	AUX
ajis02-348	131	4	illustrated	illustrate	VERB
ajis02-348	131	5	in	in	ADP
ajis02-348	131	6	figure	figure	NOUN
ajis02-348	131	7	1	1	NUM
ajis02-348	131	8	,	,	PUNCT
ajis02-348	131	9	where	where	SCONJ
ajis02-348	131	10	(	(	PUNCT
ajis02-348	131	11	2i	2i	NOUN
ajis02-348	131	12	is	be	AUX
ajis02-348	131	13	the	the	DET
ajis02-348	131	14	first	first	ADJ
ajis02-348	131	15	quartile	quartile	NOUN
ajis02-348	131	16	of	of	ADP
ajis02-348	131	17	mre	mre	PROPN
ajis02-348	131	18	and	and	CCONJ
ajis02-348	131	19	also	also	ADV
ajis02-348	131	20	expresses	express	VERB
ajis02-348	131	21	the	the	DET
ajis02-348	131	22	intersection	intersection	NOUN
ajis02-348	131	23	of	of	ADP
ajis02-348	131	24	the	the	DET
ajis02-348	131	25	curve	curve	NOUN
ajis02-348	131	26	and	and	CCONJ
ajis02-348	131	27	the	the	DET
ajis02-348	131	28	vertical	vertical	ADJ
ajis02-348	131	29	line	line	NOUN
ajis02-348	131	30	n	n	PROPN
ajis02-348	131	31	=	=	SYM
ajis02-348	131	32	nq	nq	PROPN
ajis02-348	131	33	,	,	PUNCT
ajis02-348	131	34	where	where	SCONJ
ajis02-348	131	35	nq	nq	PROPN
ajis02-348	131	36	(=	(=	ADV
ajis02-348	131	37	0.25/0	0.25/0	PROPN
ajis02-348	131	38	>	>	PUNCT
ajis02-348	131	39	s	s	VERB
ajis02-348	131	40	me	i	PRON
ajis02-348	131	41	ordinal	ordinal	ADJ
ajis02-348	131	42	number	number	NOUN
ajis02-348	131	43	of	of	ADP
ajis02-348	131	44	q\	q\	PROPN
ajis02-348	131	45	,	,	PUNCT
ajis02-348	131	46	and	and	CCONJ
ajis02-348	131	47	so	so	ADV
ajis02-348	131	48	on	on	ADV
ajis02-348	131	49	.	.	PUNCT
ajis02-348	132	1	through	through	ADP
ajis02-348	132	2	point	point	NOUN
ajis02-348	132	3	o	o	NOUN
ajis02-348	132	4	and	and	CCONJ
ajis02-348	132	5	point	point	NOUN
ajis02-348	132	6	q3	q3	PROPN
ajis02-348	132	7	a	a	DET
ajis02-348	132	8	straight	straight	ADJ
ajis02-348	132	9	line	line	NOUN
ajis02-348	132	10	is	be	AUX
ajis02-348	132	11	drawn	draw	VERB
ajis02-348	132	12	to	to	ADP
ajis02-348	132	13	intersect	intersect	NOUN
ajis02-348	132	14	with	with	ADP
ajis02-348	132	15	vertical	vertical	ADJ
ajis02-348	132	16	line	line	NOUN
ajis02-348	132	17	n	n	NOUN
ajis02-348	132	18	=	=	SYM
ajis02-348	132	19	nal	nal	ADJ
ajis02-348	132	20	point	point	NOUN
ajis02-348	132	21	q4	q4	PROPN
ajis02-348	132	22	.	.	PUNCT
ajis02-348	133	1	the	the	DET
ajis02-348	133	2	vertical	vertical	ADJ
ajis02-348	133	3	line	line	NOUN
ajis02-348	133	4	n	n	PROPN
ajis02-348	133	5	=	=	PRON
ajis02-348	133	6	nqi	nqi	NOUN
ajis02-348	133	7	meets	meet	VERB
ajis02-348	133	8	with	with	ADP
ajis02-348	133	9	straight	straight	ADJ
ajis02-348	133	10	line	line	NOUN
ajis02-348	133	11	o	o	PROPN
ajis02-348	133	12	q4	q4	PROPN
ajis02-348	133	13	at	at	ADP
ajis02-348	133	14	point	point	NOUN
ajis02-348	133	15	q	q	X
ajis02-348	133	16	}	}	PUNCT
ajis02-348	133	17	.	.	PUNCT
ajis02-348	134	1	point	point	NOUN
ajis02-348	134	2	q2	q2	PROPN
ajis02-348	134	3	is	be	AUX
ajis02-348	134	4	obtained	obtain	VERB
ajis02-348	134	5	in	in	ADP
ajis02-348	134	6	the	the	DET
ajis02-348	134	7	same	same	ADJ
ajis02-348	134	8	way	way	NOUN
ajis02-348	134	9	.	.	PUNCT
ajis02-348	135	1	ajis	ajis	ADV
ajis02-348	135	2	vol	vol	VERB
ajis02-348	135	3	5	5	NUM
ajis02-348	135	4	no	no	DET
ajis02-348	135	5	1	1	NUM
ajis02-348	135	6	september	september	NOUN
ajis02-348	135	7	1997	1997	NUM
ajis02-348	135	8	35	35	NUM
ajis02-348	135	9	mre	mre	NOUN
ajis02-348	135	10	o	o	NOUN
ajis02-348	135	11	figure	figure	NOUN
ajis02-348	135	12	1	1	NUM
ajis02-348	135	13	relationship	relationship	NOUN
ajis02-348	135	14	between	between	ADP
ajis02-348	135	15	wmq	wmq	NOUN
ajis02-348	135	16	and	and	CCONJ
ajis02-348	135	17	mre	mre	PROPN
ajis02-348	135	18	n.	n.	PROPN
ajis02-348	135	19	n	n	PROPN
ajis02-348	135	20	n	n	PROPN
ajis02-348	135	21	if	if	SCONJ
ajis02-348	135	22	there	there	PRON
ajis02-348	135	23	are	be	VERB
ajis02-348	135	24	no	no	DET
ajis02-348	135	25	outliers	outlier	NOUN
ajis02-348	135	26	with	with	ADP
ajis02-348	135	27	respect	respect	NOUN
ajis02-348	135	28	to	to	ADP
ajis02-348	135	29	mre	mre	NOUN
ajis02-348	135	30	we	we	PRON
ajis02-348	135	31	assume	assume	VERB
ajis02-348	135	32	that	that	SCONJ
ajis02-348	135	33	the	the	DET
ajis02-348	135	34	area	area	NOUN
ajis02-348	135	35	oqtq2q3q2qt	oqtq2q3q2qt	X
ajis02-348	135	36	is	be	AUX
ajis02-348	135	37	approximately	approximately	ADV
ajis02-348	135	38	equal	equal	ADJ
ajis02-348	135	39	to	to	ADP
ajis02-348	135	40	the	the	DET
ajis02-348	135	41	area	area	NOUN
ajis02-348	135	42	q^q4q4	q^q4q4	PROPN
ajis02-348	135	43	.	.	PROPN
ajis02-348	135	44	therefore	therefore	ADV
ajis02-348	135	45	the	the	DET
ajis02-348	135	46	area	area	NOUN
ajis02-348	135	47	al	al	PROPN
ajis02-348	135	48	under	under	ADP
ajis02-348	135	49	the	the	DET
ajis02-348	135	50	smoothed	smooth	VERB
ajis02-348	135	51	curve	curve	NOUN
ajis02-348	135	52	is	be	AUX
ajis02-348	135	53	approximately	approximately	ADV
ajis02-348	135	54	equal	equal	ADJ
ajis02-348	135	55	to	to	ADP
ajis02-348	135	56	the	the	DET
ajis02-348	135	57	area	area	NOUN
ajis02-348	135	58	a2	a2	PROPN
ajis02-348	135	59	of	of	ADP
ajis02-348	135	60	triangle	triangle	NOUN
ajis02-348	135	61	oq4n	oq4n	PROPN
ajis02-348	135	62	.	.	PUNCT
ajis02-348	136	1	let	let	VERB
ajis02-348	136	2	k	k	PRON
ajis02-348	136	3	be	be	AUX
ajis02-348	136	4	the	the	DET
ajis02-348	136	5	slope	slope	NOUN
ajis02-348	136	6	of	of	ADP
ajis02-348	136	7	straight	straight	ADJ
ajis02-348	136	8	line	line	NOUN
ajis02-348	136	9	o	o	PROPN
ajis02-348	136	10	q4	q4	PROPN
ajis02-348	136	11	.	.	PUNCT
ajis02-348	137	1	4mre	4mre	NUM
ajis02-348	137	2	=	=	PUNCT
ajis02-348	137	3	—	—	PUNCT
ajis02-348	137	4	n	n	CCONJ
ajis02-348	137	5	n	n	CCONJ
ajis02-348	137	6	n	n	ADV
ajis02-348	137	7	if	if	SCONJ
ajis02-348	137	8	we	we	PRON
ajis02-348	137	9	assume	assume	VERB
ajis02-348	137	10	ql	ql	X
ajis02-348	137	11	~	~	PUNCT
ajis02-348	137	12	q	q	X
ajis02-348	137	13	}	}	PUNCT
ajis02-348	137	14	and	and	CCONJ
ajis02-348	137	15	q2	q2	NOUN
ajis02-348	137	16	~	~	PUNCT
ajis02-348	137	17	q2	q2	NOUN
ajis02-348	137	18	,	,	PUNCT
ajis02-348	137	19	then	then	ADV
ajis02-348	137	20	wmq	wmq	VERB
ajis02-348	137	21	=	=	PUNCT
ajis02-348	137	22	2q2	2q2	NUM
ajis02-348	137	23	+	+	CCONJ
ajis02-348	137	24	3q3	3q3	NUM
ajis02-348	137	25	*	*	PUNCT
ajis02-348	137	26	2q2	2q2	NUM
ajis02-348	137	27	6	6	NUM
ajis02-348	137	28	025kn	025kn	NOUN
ajis02-348	137	29	+	+	X
ajis02-348	137	30	2(0.50kn	2(0.50kn	NUM
ajis02-348	137	31	)	)	PUNCT
ajis02-348	138	1	+	+	NOUN
ajis02-348	139	1	3(0.75*n	3(0.75*n	NUM
ajis02-348	139	2	)	)	PUNCT
ajis02-348	139	3	=	=	NOUN
ajis02-348	139	4	6	6	NUM
ajis02-348	139	5	in	in	ADP
ajis02-348	139	6	this	this	DET
ajis02-348	139	7	particular	particular	ADJ
ajis02-348	139	8	case	case	NOUN
ajis02-348	139	9	,	,	PUNCT
ajis02-348	139	10	as	as	SCONJ
ajis02-348	139	11	ql	ql	ADP
ajis02-348	139	12	<	<	X
ajis02-348	139	13	q	q	X
ajis02-348	139	14	}	}	PUNCT
ajis02-348	139	15	and	and	CCONJ
ajis02-348	139	16	q2	q2	NOUN
ajis02-348	139	17	<	<	X
ajis02-348	139	18	q2	q2	PROPN
ajis02-348	139	19	in	in	ADP
ajis02-348	139	20	figure	figure	NOUN
ajis02-348	139	21	1	1	NUM
ajis02-348	139	22	,	,	PUNCT
ajis02-348	139	23	wmq	wmq	X
ajis02-348	139	24	<	<	X
ajis02-348	139	25	0.58/72	0.58/72	NOUN
ajis02-348	139	26	.	.	PUNCT
ajis02-348	140	1	in	in	ADP
ajis02-348	140	2	the	the	DET
ajis02-348	140	3	situation	situation	NOUN
ajis02-348	140	4	where	where	SCONJ
ajis02-348	140	5	there	there	PRON
ajis02-348	140	6	is	be	VERB
ajis02-348	140	7	a	a	DET
ajis02-348	140	8	tendency	tendency	NOUN
ajis02-348	140	9	to	to	PART
ajis02-348	140	10	underestimate	underestimate	VERB
ajis02-348	140	11	,	,	PUNCT
ajis02-348	140	12	the	the	DET
ajis02-348	140	13	front	front	ADJ
ajis02-348	140	14	part	part	NOUN
ajis02-348	140	15	of	of	ADP
ajis02-348	140	16	the	the	DET
ajis02-348	140	17	curve	curve	NOUN
ajis02-348	140	18	goes	go	VERB
ajis02-348	140	19	up	up	ADP
ajis02-348	140	20	and	and	CCONJ
ajis02-348	140	21	the	the	DET
ajis02-348	140	22	rear	rear	ADJ
ajis02-348	140	23	part	part	NOUN
ajis02-348	140	24	comes	come	VERB
ajis02-348	140	25	down	down	ADV
ajis02-348	140	26	.	.	PUNCT
ajis02-348	141	1	hence	hence	ADV
ajis02-348	141	2	,	,	PUNCT
ajis02-348	141	3	ql	ql	X
ajis02-348	141	4	is	be	AUX
ajis02-348	141	5	close	close	ADJ
ajis02-348	141	6	to	to	ADP
ajis02-348	141	7	£	£	SYM
ajis02-348	141	8	?	?	NOUN
ajis02-348	142	1	i	i	PRON
ajis02-348	142	2	and	and	CCONJ
ajis02-348	142	3	q2	q2	NOUN
ajis02-348	142	4	is	be	AUX
ajis02-348	142	5	close	close	ADJ
ajis02-348	142	6	to	to	ADP
ajis02-348	142	7	q2	q2	NOUN
ajis02-348	142	8	.	.	PUNCT
ajis02-348	143	1	wmq	wmq	PROPN
ajis02-348	143	2	may	may	AUX
ajis02-348	143	3	be	be	AUX
ajis02-348	143	4	greater	great	ADJ
ajis02-348	143	5	than	than	ADP
ajis02-348	143	6	o.sfcn	o.sfcn	ADJ
ajis02-348	143	7	.	.	PUNCT
ajis02-348	144	1	that	that	PRON
ajis02-348	144	2	is	be	AUX
ajis02-348	144	3	wmq	wmq	PROPN
ajis02-348	144	4	>	>	X
ajis02-348	144	5	mre	mre	PROPN
ajis02-348	144	6	(	(	PUNCT
ajis02-348	144	7	the	the	DET
ajis02-348	144	8	mre	mre	PROPN
ajis02-348	144	9	favours	favour	VERB
ajis02-348	144	10	underestimation	underestimation	NOUN
ajis02-348	144	11	)	)	PUNCT
ajis02-348	144	12	.	.	PUNCT
ajis02-348	145	1	for	for	ADP
ajis02-348	145	2	an	an	DET
ajis02-348	145	3	unbiased	unbiased	ADJ
ajis02-348	145	4	estimation	estimation	NOUN
ajis02-348	145	5	,	,	PUNCT
ajis02-348	145	6	ql	ql	NOUN
ajis02-348	145	7	and	and	CCONJ
ajis02-348	145	8	q2	q2	NOUN
ajis02-348	145	9	are	be	AUX
ajis02-348	145	10	relatively	relatively	ADV
ajis02-348	145	11	further	far	ADV
ajis02-348	145	12	from	from	ADP
ajis02-348	145	13	ql	ql	NOUN
ajis02-348	145	14	and	and	CCONJ
ajis02-348	145	15	q2	q2	NOUN
ajis02-348	145	16	respectively	respectively	ADV
ajis02-348	145	17	.	.	PUNCT
ajis02-348	146	1	wmq	wmq	PROPN
ajis02-348	146	2	may	may	AUX
ajis02-348	146	3	be	be	AUX
ajis02-348	146	4	closer	close	ADJ
ajis02-348	146	5	to	to	ADP
ajis02-348	146	6	o.skn	o.skn	PROPN
ajis02-348	146	7	.	.	PUNCT
ajis02-348	147	1	that	that	PRON
ajis02-348	147	2	is	be	AUX
ajis02-348	147	3	wmq	wmq	NOUN
ajis02-348	147	4	=	=	SYM
ajis02-348	147	5	mre	mre	PROPN
ajis02-348	147	6	.	.	PROPN
ajis02-348	148	1	this	this	PRON
ajis02-348	148	2	explains	explain	VERB
ajis02-348	148	3	why	why	SCONJ
ajis02-348	148	4	wmq	wmq	NOUN
ajis02-348	148	5	and	and	CCONJ
ajis02-348	148	6	mre	mre	PROPN
ajis02-348	148	7	are	be	AUX
ajis02-348	148	8	consistent	consistent	ADJ
ajis02-348	148	9	when	when	SCONJ
ajis02-348	148	10	there	there	PRON
ajis02-348	148	11	are	be	VERB
ajis02-348	148	12	no	no	DET
ajis02-348	148	13	outliers	outlier	NOUN
ajis02-348	148	14	and	and	CCONJ
ajis02-348	148	15	the	the	DET
ajis02-348	148	16	estimation	estimation	NOUN
ajis02-348	148	17	is	be	AUX
ajis02-348	148	18	unbiased	unbiased	ADJ
ajis02-348	148	19	.	.	PUNCT
ajis02-348	149	1	measures	measure	NOUN
ajis02-348	149	2	of	of	ADP
ajis02-348	149	3	consistency	consistency	NOUN
ajis02-348	149	4	a	a	DET
ajis02-348	149	5	model	model	NOUN
ajis02-348	149	6	that	that	PRON
ajis02-348	149	7	is	be	AUX
ajis02-348	149	8	sensitive	sensitive	ADJ
ajis02-348	149	9	to	to	ADP
ajis02-348	149	10	the	the	DET
ajis02-348	149	11	influence	influence	NOUN
ajis02-348	149	12	of	of	ADP
ajis02-348	149	13	various	various	ADJ
ajis02-348	149	14	productivity	productivity	NOUN
ajis02-348	149	15	factors	factor	NOUN
ajis02-348	149	16	may	may	AUX
ajis02-348	149	17	nonetheless	nonetheless	ADV
ajis02-348	149	18	consistently	consistently	ADV
ajis02-348	149	19	overestimate	overestimate	VERB
ajis02-348	149	20	or	or	CCONJ
ajis02-348	149	21	underestimate	underestimate	ADJ
ajis02-348	149	22	development	development	NOUN
ajis02-348	149	23	,	,	PUNCT
ajis02-348	149	24	if	if	SCONJ
ajis02-348	149	25	the	the	DET
ajis02-348	149	26	standard	standard	ADJ
ajis02-348	149	27	productivity	productivity	NOUN
ajis02-348	149	28	rate	rate	NOUN
ajis02-348	149	29	assumed	assume	VERB
ajis02-348	149	30	by	by	ADP
ajis02-348	149	31	the	the	DET
ajis02-348	149	32	model	model	NOUN
ajis02-348	149	33	is	be	AUX
ajis02-348	149	34	significantly	significantly	ADV
ajis02-348	149	35	different	different	ADJ
ajis02-348	149	36	from	from	ADP
ajis02-348	149	37	that	that	PRON
ajis02-348	149	38	of	of	ADP
ajis02-348	149	39	the	the	DET
ajis02-348	149	40	environment	environment	NOUN
ajis02-348	149	41	in	in	ADP
ajis02-348	149	42	which	which	PRON
ajis02-348	149	43	the	the	DET
ajis02-348	149	44	software	software	NOUN
ajis02-348	149	45	was	be	AUX
ajis02-348	149	46	developed	develop	VERB
ajis02-348	149	47	(	(	PUNCT
ajis02-348	149	48	mukhopadhyay	mukhopadhyay	NOUN
ajis02-348	149	49	et	et	PROPN
ajis02-348	149	50	al	al	PROPN
ajis02-348	149	51	,	,	PUNCT
ajis02-348	149	52	1992	1992	NUM
ajis02-348	149	53	)	)	PUNCT
ajis02-348	149	54	.	.	PUNCT
ajis02-348	150	1	models	model	NOUN
ajis02-348	150	2	developed	develop	VERB
ajis02-348	150	3	in	in	ADP
ajis02-348	150	4	different	different	ADJ
ajis02-348	150	5	environments	environment	NOUN
ajis02-348	150	6	do	do	AUX
ajis02-348	150	7	not	not	PART
ajis02-348	150	8	work	work	VERB
ajis02-348	150	9	very	very	ADV
ajis02-348	150	10	well	well	ADV
ajis02-348	150	11	without	without	ADP
ajis02-348	150	12	calibration	calibration	NOUN
ajis02-348	150	13	.	.	PUNCT
ajis02-348	151	1	a	a	DET
ajis02-348	151	2	consistently	consistently	ADV
ajis02-348	151	3	overestimating	overestimate	VERB
ajis02-348	151	4	or	or	CCONJ
ajis02-348	151	5	underestimating	underestimating	ADJ
ajis02-348	151	6	model	model	NOUN
ajis02-348	151	7	is	be	AUX
ajis02-348	151	8	easier	easy	ADJ
ajis02-348	151	9	to	to	PART
ajis02-348	151	10	calibrate	calibrate	VERB
ajis02-348	151	11	than	than	ADP
ajis02-348	151	12	an	an	DET
ajis02-348	151	13	inconsistent	inconsistent	ADJ
ajis02-348	151	14	one	one	NUM
ajis02-348	151	15	.	.	PUNCT
ajis02-348	152	1	therefore	therefore	ADV
ajis02-348	152	2	,	,	PUNCT
ajis02-348	152	3	besides	besides	SCONJ
ajis02-348	152	4	accuracy	accuracy	NOUN
ajis02-348	152	5	,	,	PUNCT
ajis02-348	152	6	consistency	consistency	NOUN
ajis02-348	152	7	is	be	AUX
ajis02-348	152	8	another	another	DET
ajis02-348	152	9	important	important	ADJ
ajis02-348	152	10	feature	feature	NOUN
ajis02-348	152	11	for	for	ADP
ajis02-348	152	12	an	an	DET
ajis02-348	152	13	estimation	estimation	NOUN
ajis02-348	152	14	model	model	NOUN
ajis02-348	152	15	.	.	PUNCT
ajis02-348	153	1	correlation	correlation	NOUN
ajis02-348	153	2	coefficient	coefficient	NOUN
ajis02-348	153	3	of	of	ADP
ajis02-348	153	4	estimates	estimate	NOUN
ajis02-348	153	5	and	and	CCONJ
ajis02-348	153	6	actual	actual	ADJ
ajis02-348	153	7	values	value	NOUN
ajis02-348	153	8	(	(	PUNCT
ajis02-348	153	9	sdr	sdr	NOUN
ajis02-348	153	10	)	)	PUNCT
ajis02-348	153	11	to	to	PART
ajis02-348	153	12	measure	measure	VERB
ajis02-348	153	13	the	the	DET
ajis02-348	153	14	level	level	NOUN
ajis02-348	153	15	of	of	ADP
ajis02-348	153	16	consistency	consistency	NOUN
ajis02-348	153	17	,	,	PUNCT
ajis02-348	153	18	some	some	DET
ajis02-348	153	19	researchers	researcher	NOUN
ajis02-348	153	20	have	have	AUX
ajis02-348	153	21	used	use	VERB
ajis02-348	153	22	the	the	DET
ajis02-348	153	23	correlation	correlation	NOUN
ajis02-348	153	24	coefficient	coefficient	NOUN
ajis02-348	153	25	,	,	PUNCT
ajis02-348	153	26	sdr	sdr	PROPN
ajis02-348	153	27	,	,	PUNCT
ajis02-348	153	28	between	between	ADP
ajis02-348	153	29	observed	observed	ADJ
ajis02-348	153	30	and	and	CCONJ
ajis02-348	153	31	estimated	estimated	ADJ
ajis02-348	153	32	values	value	NOUN
ajis02-348	153	33	(	(	PUNCT
ajis02-348	153	34	mukhopadhyay	mukhopadhyay	NOUN
ajis02-348	153	35	et	et	PROPN
ajis02-348	153	36	al	al	PROPN
ajis02-348	153	37	,	,	PUNCT
ajis02-348	153	38	1992	1992	NUM
ajis02-348	153	39	)	)	PUNCT
ajis02-348	153	40	.	.	PUNCT
ajis02-348	154	1	this	this	DET
ajis02-348	154	2	measure	measure	NOUN
ajis02-348	154	3	tests	test	VERB
ajis02-348	154	4	the	the	DET
ajis02-348	154	5	linear	linear	PROPN
ajis02-348	154	6	association	association	NOUN
ajis02-348	154	7	between	between	ADP
ajis02-348	154	8	the	the	DET
ajis02-348	154	9	actual	actual	ADJ
ajis02-348	154	10	values	value	NOUN
ajis02-348	154	11	and	and	CCONJ
ajis02-348	154	12	estimates	estimate	NOUN
ajis02-348	154	13	.	.	PUNCT
ajis02-348	155	1	for	for	ADP
ajis02-348	155	2	a	a	DET
ajis02-348	155	3	highly	highly	ADV
ajis02-348	155	4	consistent	consistent	ADJ
ajis02-348	155	5	model	model	NOUN
ajis02-348	155	6	,	,	PUNCT
ajis02-348	155	7	r	r	NOUN
ajis02-348	155	8	should	should	AUX
ajis02-348	155	9	be	be	AUX
ajis02-348	155	10	close	close	ADJ
ajis02-348	155	11	to	to	ADP
ajis02-348	155	12	1	1	NUM
ajis02-348	155	13	(	(	PUNCT
ajis02-348	155	14	-!</	-!</	PROPN
ajis02-348	155	15	?	?	PUNCT
ajis02-348	155	16	<	<	X
ajis02-348	155	17	1	1	NUM
ajis02-348	155	18	)	)	PUNCT
ajis02-348	155	19	,	,	PUNCT
ajis02-348	155	20	otherwise	otherwise	ADV
ajis02-348	155	21	it	it	PRON
ajis02-348	155	22	is	be	AUX
ajis02-348	155	23	ajis	ajis	ADV
ajis02-348	155	24	vol	vol	NOUN
ajis02-348	155	25	5	5	NUM
ajis02-348	155	26	no	no	DET
ajis02-348	155	27	1	1	NUM
ajis02-348	155	28	september	september	NOUN
ajis02-348	155	29	1997	1997	NUM
ajis02-348	155	30	36	36	NUM
ajis02-348	155	31	close	close	ADV
ajis02-348	155	32	to	to	ADP
ajis02-348	155	33	0	0	NUM
ajis02-348	155	34	.	.	PUNCT
ajis02-348	156	1	if	if	SCONJ
ajis02-348	156	2	r	r	NOUN
ajis02-348	156	3	is	be	AUX
ajis02-348	156	4	negative	negative	ADJ
ajis02-348	156	5	,	,	PUNCT
ajis02-348	156	6	it	it	PRON
ajis02-348	156	7	indicates	indicate	VERB
ajis02-348	156	8	that	that	SCONJ
ajis02-348	156	9	larger	large	ADJ
ajis02-348	156	10	actual	actual	ADJ
ajis02-348	156	11	values	value	NOUN
ajis02-348	156	12	are	be	AUX
ajis02-348	156	13	associated	associate	VERB
ajis02-348	156	14	with	with	ADP
ajis02-348	156	15	smaller	small	ADJ
ajis02-348	156	16	estimates	estimate	NOUN
ajis02-348	156	17	.	.	PUNCT
ajis02-348	157	1	r	r	NOUN
ajis02-348	157	2	is	be	AUX
ajis02-348	157	3	the	the	DET
ajis02-348	157	4	square	square	ADJ
ajis02-348	157	5	root	root	NOUN
ajis02-348	157	6	of	of	ADP
ajis02-348	157	7	r	r	NOUN
ajis02-348	157	8	,	,	PUNCT
ajis02-348	157	9	the	the	DET
ajis02-348	157	10	coefficient	coefficient	NOUN
ajis02-348	157	11	of	of	ADP
ajis02-348	157	12	determination	determination	NOUN
ajis02-348	157	13	introduced	introduce	VERB
ajis02-348	157	14	earlier	early	ADV
ajis02-348	157	15	.	.	PUNCT
ajis02-348	158	1	so	so	ADV
ajis02-348	158	2	sdr	sdr	PROPN
ajis02-348	158	3	is	be	AUX
ajis02-348	158	4	not	not	PART
ajis02-348	158	5	consistent	consistent	ADJ
ajis02-348	158	6	with	with	ADP
ajis02-348	158	7	the	the	DET
ajis02-348	158	8	ratio	ratio	NOUN
ajis02-348	158	9	measure	measure	NOUN
ajis02-348	158	10	which	which	PRON
ajis02-348	158	11	has	have	AUX
ajis02-348	158	12	been	be	AUX
ajis02-348	158	13	illustrated	illustrate	VERB
ajis02-348	158	14	to	to	PART
ajis02-348	158	15	be	be	AUX
ajis02-348	158	16	more	more	ADV
ajis02-348	158	17	suitable	suitable	ADJ
ajis02-348	158	18	for	for	ADP
ajis02-348	158	19	software	software	NOUN
ajis02-348	158	20	cost	cost	NOUN
ajis02-348	158	21	estimation	estimation	NOUN
ajis02-348	158	22	.	.	PUNCT
ajis02-348	159	1	moreover	moreover	ADV
ajis02-348	159	2	,	,	PUNCT
ajis02-348	159	3	as	as	ADP
ajis02-348	159	4	r2	r2	PROPN
ajis02-348	159	5	,	,	PUNCT
ajis02-348	159	6	sdr	sdr	PROPN
ajis02-348	159	7	is	be	AUX
ajis02-348	159	8	influenced	influence	VERB
ajis02-348	159	9	by	by	ADP
ajis02-348	159	10	the	the	DET
ajis02-348	159	11	variance	variance	NOUN
ajis02-348	159	12	of	of	ADP
ajis02-348	159	13	data	datum	NOUN
ajis02-348	159	14	.	.	PUNCT
ajis02-348	160	1	the	the	PRON
ajis02-348	160	2	greater	great	ADJ
ajis02-348	160	3	the	the	DET
ajis02-348	160	4	variance	variance	NOUN
ajis02-348	160	5	of	of	ADP
ajis02-348	160	6	actual	actual	ADJ
ajis02-348	160	7	values	value	NOUN
ajis02-348	160	8	,	,	PUNCT
ajis02-348	160	9	the	the	PRON
ajis02-348	160	10	larger	large	ADJ
ajis02-348	160	11	the	the	DET
ajis02-348	160	12	denominator	denominator	NOUN
ajis02-348	160	13	in	in	ADP
ajis02-348	160	14	the	the	DET
ajis02-348	160	15	expression	expression	NOUN
ajis02-348	160	16	.	.	PUNCT
ajis02-348	161	1	r	r	NOUN
ajis02-348	161	2	varies	vary	VERB
ajis02-348	161	3	according	accord	VERB
ajis02-348	161	4	to	to	ADP
ajis02-348	161	5	not	not	PART
ajis02-348	161	6	only	only	ADV
ajis02-348	161	7	the	the	DET
ajis02-348	161	8	estimation	estimation	NOUN
ajis02-348	161	9	accuracy	accuracy	NOUN
ajis02-348	161	10	but	but	CCONJ
ajis02-348	161	11	also	also	ADV
ajis02-348	161	12	the	the	DET
ajis02-348	161	13	variance	variance	NOUN
ajis02-348	161	14	of	of	ADP
ajis02-348	161	15	data	datum	NOUN
ajis02-348	161	16	.	.	PUNCT
ajis02-348	162	1	we	we	PRON
ajis02-348	162	2	need	need	VERB
ajis02-348	162	3	a	a	DET
ajis02-348	162	4	consistency	consistency	NOUN
ajis02-348	162	5	measure	measure	NOUN
ajis02-348	162	6	,	,	PUNCT
ajis02-348	162	7	which	which	PRON
ajis02-348	162	8	assesses	assess	VERB
ajis02-348	162	9	the	the	DET
ajis02-348	162	10	estimation	estimation	NOUN
ajis02-348	162	11	on	on	ADP
ajis02-348	162	12	the	the	DET
ajis02-348	162	13	basis	basis	NOUN
ajis02-348	162	14	of	of	ADP
ajis02-348	162	15	ratio	ratio	NOUN
ajis02-348	162	16	measure	measure	NOUN
ajis02-348	162	17	and	and	CCONJ
ajis02-348	162	18	is	be	AUX
ajis02-348	162	19	independent	independent	ADJ
ajis02-348	162	20	to	to	ADP
ajis02-348	162	21	the	the	DET
ajis02-348	162	22	distribution	distribution	NOUN
ajis02-348	162	23	of	of	ADP
ajis02-348	162	24	the	the	DET
ajis02-348	162	25	actual	actual	ADJ
ajis02-348	162	26	observations	observation	NOUN
ajis02-348	162	27	.	.	PUNCT
ajis02-348	163	1	a	a	DET
ajis02-348	163	2	new	new	ADJ
ajis02-348	163	3	measure	measure	NOUN
ajis02-348	163	4	of	of	ADP
ajis02-348	163	5	consistency	consistency	NOUN
ajis02-348	163	6	suppose	suppose	VERB
ajis02-348	163	7	that	that	SCONJ
ajis02-348	163	8	a	a	DET
ajis02-348	163	9	model	model	NOUN
ajis02-348	163	10	was	be	AUX
ajis02-348	163	11	developed	develop	VERB
ajis02-348	163	12	in	in	ADP
ajis02-348	163	13	environment	environment	NOUN
ajis02-348	163	14	a	a	PRON
ajis02-348	163	15	and	and	CCONJ
ajis02-348	163	16	a	a	DET
ajis02-348	163	17	set	set	NOUN
ajis02-348	163	18	of	of	ADP
ajis02-348	163	19	data	datum	NOUN
ajis02-348	163	20	,	,	PUNCT
ajis02-348	163	21	which	which	PRON
ajis02-348	163	22	was	be	AUX
ajis02-348	163	23	collected	collect	VERB
ajis02-348	163	24	from	from	ADP
ajis02-348	163	25	software	software	NOUN
ajis02-348	163	26	projects	project	NOUN
ajis02-348	163	27	developed	develop	VERB
ajis02-348	163	28	in	in	ADP
ajis02-348	163	29	environment	environment	NOUN
ajis02-348	163	30	b	b	NOUN
ajis02-348	163	31	,	,	PUNCT
ajis02-348	163	32	is	be	AUX
ajis02-348	163	33	used	use	VERB
ajis02-348	163	34	to	to	PART
ajis02-348	163	35	test	test	VERB
ajis02-348	163	36	the	the	DET
ajis02-348	163	37	estimation	estimation	NOUN
ajis02-348	163	38	consistency	consistency	NOUN
ajis02-348	163	39	of	of	ADP
ajis02-348	163	40	the	the	DET
ajis02-348	163	41	model	model	NOUN
ajis02-348	163	42	.	.	PUNCT
ajis02-348	164	1	for	for	ADP
ajis02-348	164	2	each	each	DET
ajis02-348	164	3	estimation	estimation	NOUN
ajis02-348	164	4	there	there	PRON
ajis02-348	164	5	is	be	VERB
ajis02-348	164	6	a	a	DET
ajis02-348	164	7	ratio	ratio	NOUN
ajis02-348	164	8	est	est	NOUN
ajis02-348	164	9	.	.	PUNCT
ajis02-348	165	1	(	(	PUNCT
ajis02-348	165	2	i	i	NOUN
ajis02-348	165	3	=	=	VERB
ajis02-348	165	4	1.2,3	1.2,3	X
ajis02-348	165	5	.....	.....	PUNCT
ajis02-348	165	6	n	n	CCONJ
ajis02-348	165	7	)	)	PUNCT
ajis02-348	165	8	acti	acti	VERB
ajis02-348	165	9	in	in	ADP
ajis02-348	165	10	the	the	DET
ajis02-348	165	11	case	case	NOUN
ajis02-348	165	12	of	of	ADP
ajis02-348	165	13	consistent	consistent	ADJ
ajis02-348	165	14	estimation	estimation	NOUN
ajis02-348	165	15	,	,	PUNCT
ajis02-348	165	16	the	the	DET
ajis02-348	165	17	values	value	NOUN
ajis02-348	165	18	of	of	ADP
ajis02-348	165	19	r.	r.	PROPN
ajis02-348	165	20	(	(	PUNCT
ajis02-348	165	21	i	i	NOUN
ajis02-348	165	22	=	=	PUNCT
ajis02-348	165	23	l,2,3,	l,2,3,	PROPN
ajis02-348	165	24	...	...	PUNCT
ajis02-348	165	25	,/i	,/i	PRON
ajis02-348	165	26	)	)	PUNCT
ajis02-348	165	27	are	be	AUX
ajis02-348	165	28	close	close	ADJ
ajis02-348	165	29	to	to	ADP
ajis02-348	165	30	one	one	NUM
ajis02-348	165	31	another	another	DET
ajis02-348	165	32	.	.	PUNCT
ajis02-348	166	1	on	on	ADP
ajis02-348	166	2	the	the	DET
ajis02-348	166	3	other	other	ADJ
ajis02-348	166	4	hand	hand	NOUN
ajis02-348	166	5	,	,	PUNCT
ajis02-348	166	6	if	if	SCONJ
ajis02-348	166	7	the	the	DET
ajis02-348	166	8	values	value	NOUN
ajis02-348	166	9	of	of	ADP
ajis02-348	166	10	r	r	NOUN
ajis02-348	166	11	;	;	PUNCT
ajis02-348	166	12	spread	spread	VERB
ajis02-348	166	13	over	over	ADP
ajis02-348	166	14	a	a	DET
ajis02-348	166	15	wide	wide	ADJ
ajis02-348	166	16	range	range	NOUN
ajis02-348	166	17	,	,	PUNCT
ajis02-348	166	18	the	the	DET
ajis02-348	166	19	estimation	estimation	NOUN
ajis02-348	166	20	is	be	AUX
ajis02-348	166	21	not	not	PART
ajis02-348	166	22	consistent	consistent	ADJ
ajis02-348	166	23	.	.	PUNCT
ajis02-348	167	1	the	the	PRON
ajis02-348	167	2	closer	close	ADV
ajis02-348	167	3	to	to	ADP
ajis02-348	167	4	one	one	NUM
ajis02-348	167	5	another	another	DET
ajis02-348	167	6	the	the	DET
ajis02-348	167	7	values	value	NOUN
ajis02-348	167	8	of	of	ADP
ajis02-348	167	9	rf	rf	NOUN
ajis02-348	167	10	are	be	AUX
ajis02-348	167	11	,	,	PUNCT
ajis02-348	167	12	the	the	PRON
ajis02-348	167	13	more	more	ADV
ajis02-348	167	14	consistent	consistent	ADJ
ajis02-348	167	15	the	the	DET
ajis02-348	167	16	estimation	estimation	NOUN
ajis02-348	167	17	is	be	AUX
ajis02-348	167	18	.	.	PUNCT
ajis02-348	168	1	statistically	statistically	ADV
ajis02-348	168	2	,	,	PUNCT
ajis02-348	168	3	standard	standard	ADJ
ajis02-348	168	4	deviation	deviation	NOUN
ajis02-348	168	5	is	be	AUX
ajis02-348	168	6	a	a	DET
ajis02-348	168	7	measure	measure	NOUN
ajis02-348	168	8	of	of	ADP
ajis02-348	168	9	variation	variation	NOUN
ajis02-348	168	10	or	or	CCONJ
ajis02-348	168	11	spread	spread	NOUN
ajis02-348	168	12	of	of	ADP
ajis02-348	168	13	the	the	DET
ajis02-348	168	14	ri	ri	PROPN
ajis02-348	168	15	'	'	PUNCT
ajis02-348	168	16	s.	s.	PROPN
ajis02-348	169	1	so	so	ADV
ajis02-348	169	2	it	it	PRON
ajis02-348	169	3	is	be	AUX
ajis02-348	169	4	proposed	propose	VERB
ajis02-348	169	5	to	to	PART
ajis02-348	169	6	use	use	VERB
ajis02-348	169	7	the	the	DET
ajis02-348	169	8	standard	standard	ADJ
ajis02-348	169	9	deviation	deviation	NOUN
ajis02-348	169	10	of	of	ADP
ajis02-348	169	11	r	r	NOUN
ajis02-348	169	12	;	;	PUNCT
ajis02-348	169	13	(	(	PUNCT
ajis02-348	169	14	sdr	sdr	NOUN
ajis02-348	169	15	)	)	PUNCT
ajis02-348	169	16	as	as	ADP
ajis02-348	169	17	a	a	DET
ajis02-348	169	18	measure	measure	NOUN
ajis02-348	169	19	of	of	ADP
ajis02-348	169	20	estimation	estimation	NOUN
ajis02-348	169	21	consistency	consistency	NOUN
ajis02-348	169	22	,	,	PUNCT
ajis02-348	170	1	sdr	sdr	PROPN
ajis02-348	170	2	=	=	SYM
ajis02-348	170	3	n-\	n-\	PROPN
ajis02-348	170	4	where	where	SCONJ
ajis02-348	170	5	r	r	NOUN
ajis02-348	170	6	is	be	AUX
ajis02-348	170	7	the	the	DET
ajis02-348	170	8	mean	mean	NOUN
ajis02-348	170	9	of	of	ADP
ajis02-348	170	10	r.	r.	PROPN
ajis02-348	170	11	's	's	PART
ajis02-348	170	12	(;	(;	NOUN
ajis02-348	170	13	=	=	NOUN
ajis02-348	170	14	l,2,3,	l,2,3,	NUM
ajis02-348	170	15	...	...	PUNCT
ajis02-348	170	16	,/i	,/i	PRON
ajis02-348	170	17	)	)	PUNCT
ajis02-348	170	18	the	the	PRON
ajis02-348	170	19	smaller	small	ADJ
ajis02-348	170	20	the	the	DET
ajis02-348	170	21	sdr	sdr	PROPN
ajis02-348	170	22	,	,	PUNCT
ajis02-348	170	23	the	the	PRON
ajis02-348	170	24	more	more	ADV
ajis02-348	170	25	consistent	consistent	ADJ
ajis02-348	170	26	the	the	DET
ajis02-348	170	27	estimation	estimation	NOUN
ajis02-348	170	28	.	.	PUNCT
ajis02-348	171	1	it	it	PRON
ajis02-348	171	2	can	can	AUX
ajis02-348	171	3	be	be	AUX
ajis02-348	171	4	shown	show	VERB
ajis02-348	171	5	that	that	SCONJ
ajis02-348	171	6	standard	standard	ADJ
ajis02-348	171	7	deviation	deviation	NOUN
ajis02-348	171	8	of	of	ADP
ajis02-348	171	9	relative	relative	ADJ
ajis02-348	171	10	error	error	NOUN
ajis02-348	171	11	is	be	AUX
ajis02-348	171	12	equal	equal	ADJ
ajis02-348	171	13	to	to	ADP
ajis02-348	171	14	sdr	sdr	PROPN
ajis02-348	171	15	.	.	PUNCT
ajis02-348	172	1	because	because	SCONJ
ajis02-348	172	2	the	the	DET
ajis02-348	172	3	assessment	assessment	NOUN
ajis02-348	172	4	of	of	ADP
ajis02-348	172	5	estimation	estimation	NOUN
ajis02-348	172	6	accuracy	accuracy	NOUN
ajis02-348	172	7	is	be	AUX
ajis02-348	172	8	based	base	VERB
ajis02-348	172	9	on	on	ADP
ajis02-348	172	10	relative	relative	ADJ
ajis02-348	172	11	error	error	NOUN
ajis02-348	172	12	,	,	PUNCT
ajis02-348	172	13	sdr	sdr	PROPN
ajis02-348	172	14	is	be	AUX
ajis02-348	172	15	related	relate	VERB
ajis02-348	172	16	to	to	ADP
ajis02-348	172	17	estimation	estimation	NOUN
ajis02-348	172	18	accuracy	accuracy	NOUN
ajis02-348	172	19	.	.	PUNCT
ajis02-348	173	1	therefore	therefore	ADV
ajis02-348	173	2	sdr	sdr	PROPN
ajis02-348	173	3	can	can	AUX
ajis02-348	173	4	be	be	AUX
ajis02-348	173	5	used	use	VERB
ajis02-348	173	6	to	to	PART
ajis02-348	173	7	calibrate	calibrate	VERB
ajis02-348	173	8	a	a	DET
ajis02-348	173	9	model	model	NOUN
ajis02-348	173	10	in	in	ADP
ajis02-348	173	11	order	order	NOUN
ajis02-348	173	12	to	to	PART
ajis02-348	173	13	improve	improve	VERB
ajis02-348	173	14	the	the	DET
ajis02-348	173	15	estimation	estimation	NOUN
ajis02-348	173	16	accuracy	accuracy	NOUN
ajis02-348	173	17	in	in	ADP
ajis02-348	173	18	different	different	ADJ
ajis02-348	173	19	environments	environment	NOUN
ajis02-348	173	20	.	.	PUNCT
ajis02-348	174	1	regression	regression	NOUN
ajis02-348	174	2	criteria	criterion	NOUN
ajis02-348	174	3	in	in	ADP
ajis02-348	174	4	a	a	DET
ajis02-348	174	5	previous	previous	ADJ
ajis02-348	174	6	section	section	NOUN
ajis02-348	174	7	,	,	PUNCT
ajis02-348	174	8	it	it	PRON
ajis02-348	174	9	is	be	AUX
ajis02-348	174	10	argued	argue	VERB
ajis02-348	174	11	that	that	SCONJ
ajis02-348	174	12	ratio	ratio	NOUN
ajis02-348	174	13	measures	measure	NOUN
ajis02-348	174	14	are	be	AUX
ajis02-348	174	15	more	more	ADV
ajis02-348	174	16	suitable	suitable	ADJ
ajis02-348	174	17	for	for	ADP
ajis02-348	174	18	accuracy	accuracy	NOUN
ajis02-348	174	19	assessment	assessment	NOUN
ajis02-348	174	20	.	.	PUNCT
ajis02-348	175	1	we	we	PRON
ajis02-348	175	2	now	now	ADV
ajis02-348	175	3	consider	consider	VERB
ajis02-348	175	4	how	how	SCONJ
ajis02-348	175	5	to	to	PART
ajis02-348	175	6	determine	determine	VERB
ajis02-348	175	7	model	model	NOUN
ajis02-348	175	8	parameters	parameter	NOUN
ajis02-348	175	9	so	so	SCONJ
ajis02-348	175	10	as	as	SCONJ
ajis02-348	175	11	to	to	PART
ajis02-348	175	12	satisfy	satisfy	VERB
ajis02-348	175	13	the	the	DET
ajis02-348	175	14	ratio	ratio	NOUN
ajis02-348	175	15	measures	measure	NOUN
ajis02-348	175	16	of	of	ADP
ajis02-348	175	17	accuracy	accuracy	NOUN
ajis02-348	175	18	.	.	PUNCT
ajis02-348	176	1	in	in	ADP
ajis02-348	176	2	modeling	modeling	NOUN
ajis02-348	176	3	costing	cost	VERB
ajis02-348	176	4	formulas	formula	NOUN
ajis02-348	176	5	,	,	PUNCT
ajis02-348	176	6	regression	regression	NOUN
ajis02-348	176	7	is	be	AUX
ajis02-348	176	8	the	the	DET
ajis02-348	176	9	basic	basic	ADJ
ajis02-348	176	10	technique	technique	NOUN
ajis02-348	176	11	to	to	PART
ajis02-348	176	12	determine	determine	VERB
ajis02-348	176	13	model	model	NOUN
ajis02-348	176	14	parameters	parameter	NOUN
ajis02-348	176	15	.	.	PUNCT
ajis02-348	177	1	the	the	DET
ajis02-348	177	2	traditional	traditional	ADJ
ajis02-348	177	3	method	method	NOUN
ajis02-348	177	4	employed	employ	VERB
ajis02-348	177	5	by	by	ADP
ajis02-348	177	6	most	most	ADJ
ajis02-348	177	7	statistical	statistical	ADJ
ajis02-348	177	8	software	software	NOUN
ajis02-348	177	9	is	be	AUX
ajis02-348	177	10	least	least	ADJ
ajis02-348	177	11	squares	square	NOUN
ajis02-348	177	12	(	(	PUNCT
ajis02-348	177	13	ls	ls	PROPN
ajis02-348	177	14	)	)	PUNCT
ajis02-348	177	15	(	(	PUNCT
ajis02-348	177	16	khoshgoftaar	khoshgoftaar	NOUN
ajis02-348	177	17	et	et	PROPN
ajis02-348	177	18	al	al	PROPN
ajis02-348	177	19	,	,	PUNCT
ajis02-348	177	20	1992	1992	NUM
ajis02-348	177	21	)	)	PUNCT
ajis02-348	177	22	,	,	PUNCT
ajis02-348	177	23	which	which	PRON
ajis02-348	177	24	can	can	AUX
ajis02-348	177	25	be	be	AUX
ajis02-348	177	26	expressed	express	VERB
ajis02-348	177	27	as	as	ADP
ajis02-348	177	28	min	min	NOUN
ajis02-348	177	29	[	[	PUNCT
ajis02-348	177	30	this	this	PRON
ajis02-348	177	31	is	be	AUX
ajis02-348	177	32	a	a	DET
ajis02-348	177	33	criterion	criterion	NOUN
ajis02-348	177	34	of	of	ADP
ajis02-348	177	35	difference	difference	NOUN
ajis02-348	177	36	measure	measure	NOUN
ajis02-348	177	37	.	.	PUNCT
ajis02-348	178	1	this	this	DET
ajis02-348	178	2	criterion	criterion	NOUN
ajis02-348	178	3	can	can	AUX
ajis02-348	178	4	not	not	PART
ajis02-348	178	5	lead	lead	VERB
ajis02-348	178	6	to	to	ADP
ajis02-348	178	7	an	an	DET
ajis02-348	178	8	optimised	optimise	VERB
ajis02-348	178	9	functional	functional	ADJ
ajis02-348	178	10	expression	expression	NOUN
ajis02-348	178	11	of	of	ADP
ajis02-348	178	12	the	the	DET
ajis02-348	178	13	model	model	NOUN
ajis02-348	178	14	as	as	SCONJ
ajis02-348	178	15	ratio	ratio	NOUN
ajis02-348	178	16	measure	measure	NOUN
ajis02-348	178	17	should	should	AUX
ajis02-348	178	18	be	be	AUX
ajis02-348	178	19	used	use	VERB
ajis02-348	178	20	to	to	PART
ajis02-348	178	21	assess	assess	VERB
ajis02-348	178	22	the	the	DET
ajis02-348	178	23	prediction	prediction	NOUN
ajis02-348	178	24	accuracy	accuracy	NOUN
ajis02-348	178	25	.	.	PUNCT
ajis02-348	179	1	in	in	ADP
ajis02-348	179	2	modelling	model	VERB
ajis02-348	179	3	formulas	formula	NOUN
ajis02-348	179	4	for	for	ADP
ajis02-348	179	5	software	software	NOUN
ajis02-348	179	6	cost	cost	NOUN
ajis02-348	179	7	estimation	estimation	NOUN
ajis02-348	179	8	,	,	PUNCT
ajis02-348	179	9	the	the	DET
ajis02-348	179	10	criterion	criterion	NOUN
ajis02-348	179	11	of	of	ADP
ajis02-348	179	12	minimising	minimise	VERB
ajis02-348	179	13	the	the	DET
ajis02-348	179	14	sum	sum	NOUN
ajis02-348	179	15	of	of	ADP
ajis02-348	179	16	squared	squared	ADJ
ajis02-348	179	17	relative	relative	ADJ
ajis02-348	179	18	errors	error	NOUN
ajis02-348	179	19	has	have	AUX
ajis02-348	179	20	been	be	AUX
ajis02-348	179	21	used	use	VERB
ajis02-348	179	22	(	(	PUNCT
ajis02-348	179	23	conte	conte	PROPN
ajis02-348	179	24	et	et	PROPN
ajis02-348	179	25	al	al	PROPN
ajis02-348	179	26	,	,	PUNCT
ajis02-348	179	27	1986	1986	NUM
ajis02-348	179	28	)	)	PUNCT
ajis02-348	179	29	.	.	PUNCT
ajis02-348	180	1	this	this	DET
ajis02-348	180	2	criterion	criterion	NOUN
ajis02-348	180	3	is	be	AUX
ajis02-348	180	4	not	not	PART
ajis02-348	180	5	suitable	suitable	ADJ
ajis02-348	180	6	either	either	ADV
ajis02-348	180	7	.	.	PUNCT
ajis02-348	181	1	in	in	ADP
ajis02-348	181	2	the	the	DET
ajis02-348	181	3	expression	expression	NOUN
ajis02-348	181	4	of	of	ADP
ajis02-348	181	5	mre	mre	PROPN
ajis02-348	181	6	ajis	ajis	ADV
ajis02-348	181	7	vol	vol	VERB
ajis02-348	181	8	5	5	NUM
ajis02-348	181	9	no	no	DET
ajis02-348	181	10	1	1	NUM
ajis02-348	181	11	september	september	NOUN
ajis02-348	181	12	1997	1997	NUM
ajis02-348	181	13	37	37	NUM
ajis02-348	181	14	mrei	mrei	NOUN
ajis02-348	181	15	=	=	SYM
ajis02-348	181	16	act	act	PROPN
ajis02-348	181	17	,	,	PUNCT
ajis02-348	181	18	est	est	X
ajis02-348	181	19	.	.	PUNCT
ajis02-348	182	1	act	act	PROPN
ajis02-348	182	2	:	:	PUNCT
ajis02-348	182	3	1est	1est	NUM
ajis02-348	182	4	:	:	PUNCT
ajis02-348	182	5	act	act	PROPN
ajis02-348	182	6	.	.	PUNCT
ajis02-348	183	1	actf	actf	ADJ
ajis02-348	183	2	is	be	AUX
ajis02-348	183	3	positive	positive	ADJ
ajis02-348	183	4	for	for	ADP
ajis02-348	183	5	all	all	DET
ajis02-348	183	6	i	i	PRON
ajis02-348	183	7	,	,	PUNCT
ajis02-348	183	8	and	and	CCONJ
ajis02-348	183	9	est	est	X
ajis02-348	183	10	(	(	PUNCT
ajis02-348	183	11	should	should	AUX
ajis02-348	183	12	be	be	AUX
ajis02-348	183	13	positive	positive	ADJ
ajis02-348	183	14	for	for	ADP
ajis02-348	183	15	all	all	DET
ajis02-348	183	16	i.	i.	NOUN
ajis02-348	183	17	in	in	ADP
ajis02-348	183	18	the	the	DET
ajis02-348	183	19	case	case	NOUN
ajis02-348	183	20	of	of	ADP
ajis02-348	183	21	overestimation	overestimation	NOUN
ajis02-348	183	22	,	,	PUNCT
ajis02-348	183	23	that	that	PRON
ajis02-348	183	24	is	be	AUX
ajis02-348	183	25	est	est	PROPN
ajis02-348	183	26	>	>	X
ajis02-348	183	27	act	act	PROPN
ajis02-348	183	28	,	,	PUNCT
ajis02-348	183	29	an	an	DET
ajis02-348	183	30	mre	mre	NOUN
ajis02-348	183	31	can	can	AUX
ajis02-348	183	32	be	be	AUX
ajis02-348	183	33	greater	great	ADJ
ajis02-348	183	34	than	than	ADP
ajis02-348	183	35	1	1	NUM
ajis02-348	183	36	,	,	PUNCT
ajis02-348	183	37	but	but	CCONJ
ajis02-348	183	38	in	in	ADP
ajis02-348	183	39	the	the	DET
ajis02-348	183	40	case	case	NOUN
ajis02-348	183	41	of	of	ADP
ajis02-348	183	42	underestimation	underestimation	NOUN
ajis02-348	183	43	,	,	PUNCT
ajis02-348	183	44	that	that	PRON
ajis02-348	183	45	is	be	AUX
ajis02-348	183	46	estf	estf	NOUN
ajis02-348	183	47	<	<	X
ajis02-348	183	48	act	act	PROPN
ajis02-348	183	49	(	(	PUNCT
ajis02-348	183	50	,	,	PUNCT
ajis02-348	183	51	an	an	DET
ajis02-348	183	52	mre	mre	NOUN
ajis02-348	183	53	can	can	AUX
ajis02-348	183	54	never	never	ADV
ajis02-348	183	55	be	be	AUX
ajis02-348	183	56	greater	great	ADJ
ajis02-348	183	57	than	than	ADP
ajis02-348	183	58	1	1	NUM
ajis02-348	183	59	.	.	PUNCT
ajis02-348	184	1	this	this	PRON
ajis02-348	184	2	explains	explain	VERB
ajis02-348	184	3	the	the	DET
ajis02-348	184	4	problem	problem	NOUN
ajis02-348	184	5	discussed	discuss	VERB
ajis02-348	184	6	in	in	ADP
ajis02-348	184	7	a	a	DET
ajis02-348	184	8	previous	previous	ADJ
ajis02-348	184	9	section	section	NOUN
ajis02-348	184	10	that	that	SCONJ
ajis02-348	184	11	mre	mre	PROPN
ajis02-348	184	12	favours	favour	VERB
ajis02-348	184	13	underestimation	underestimation	NOUN
ajis02-348	184	14	if	if	SCONJ
ajis02-348	184	15	it	it	PRON
ajis02-348	184	16	is	be	AUX
ajis02-348	184	17	used	use	VERB
ajis02-348	184	18	as	as	ADP
ajis02-348	184	19	a	a	DET
ajis02-348	184	20	measure	measure	NOUN
ajis02-348	184	21	of	of	ADP
ajis02-348	184	22	accuracy	accuracy	NOUN
ajis02-348	184	23	.	.	PUNCT
ajis02-348	185	1	to	to	PART
ajis02-348	185	2	achieve	achieve	VERB
ajis02-348	185	3	the	the	DET
ajis02-348	185	4	least	least	ADJ
ajis02-348	185	5	sum	sum	NOUN
ajis02-348	185	6	of	of	ADP
ajis02-348	185	7	squared	squared	ADJ
ajis02-348	185	8	relative	relative	ADJ
ajis02-348	185	9	errors	error	NOUN
ajis02-348	185	10	,	,	PUNCT
ajis02-348	185	11	the	the	DET
ajis02-348	185	12	regression	regression	NOUN
ajis02-348	185	13	technique	technique	NOUN
ajis02-348	185	14	makes	make	VERB
ajis02-348	185	15	as	as	ADP
ajis02-348	185	16	many	many	ADJ
ajis02-348	185	17	negative	negative	ADJ
ajis02-348	185	18	relative	relative	ADJ
ajis02-348	185	19	errors	error	NOUN
ajis02-348	185	20	as	as	ADP
ajis02-348	185	21	possible	possible	ADJ
ajis02-348	185	22	.	.	PUNCT
ajis02-348	186	1	therefore	therefore	ADV
ajis02-348	186	2	,	,	PUNCT
ajis02-348	186	3	if	if	SCONJ
ajis02-348	186	4	least	least	ADJ
ajis02-348	186	5	sum	sum	NOUN
ajis02-348	186	6	of	of	ADP
ajis02-348	186	7	squared	squared	ADJ
ajis02-348	186	8	relative	relative	ADJ
ajis02-348	186	9	errors	error	NOUN
ajis02-348	186	10	is	be	AUX
ajis02-348	186	11	used	use	VERB
ajis02-348	186	12	as	as	ADP
ajis02-348	186	13	the	the	DET
ajis02-348	186	14	regression	regression	NOUN
ajis02-348	186	15	criterion	criterion	NOUN
ajis02-348	186	16	,	,	PUNCT
ajis02-348	186	17	the	the	DET
ajis02-348	186	18	obtained	obtain	VERB
ajis02-348	186	19	formula	formula	NOUN
ajis02-348	186	20	will	will	AUX
ajis02-348	186	21	be	be	AUX
ajis02-348	186	22	one	one	NUM
ajis02-348	186	23	that	that	PRON
ajis02-348	186	24	systematically	systematically	ADV
ajis02-348	186	25	underestimates	underestimate	VERB
ajis02-348	186	26	the	the	DET
ajis02-348	186	27	effort	effort	NOUN
ajis02-348	186	28	.	.	PUNCT
ajis02-348	187	1	it	it	PRON
ajis02-348	187	2	has	have	AUX
ajis02-348	187	3	been	be	AUX
ajis02-348	187	4	found	find	VERB
ajis02-348	187	5	in	in	ADP
ajis02-348	187	6	practice	practice	NOUN
ajis02-348	187	7	that	that	SCONJ
ajis02-348	187	8	in	in	ADP
ajis02-348	187	9	most	most	ADJ
ajis02-348	187	10	cases	case	NOUN
ajis02-348	187	11	of	of	ADP
ajis02-348	187	12	software	software	NOUN
ajis02-348	187	13	cost	cost	NOUN
ajis02-348	187	14	estimation	estimation	NOUN
ajis02-348	187	15	,	,	PUNCT
ajis02-348	187	16	the	the	DET
ajis02-348	187	17	errors	error	NOUN
ajis02-348	187	18	came	come	VERB
ajis02-348	187	19	from	from	ADP
ajis02-348	187	20	underestimation	underestimation	NOUN
ajis02-348	187	21	(	(	PUNCT
ajis02-348	187	22	lederer	lederer	PROPN
ajis02-348	187	23	et	et	PROPN
ajis02-348	187	24	al	al	PROPN
ajis02-348	187	25	,	,	PUNCT
ajis02-348	187	26	1993	1993	NUM
ajis02-348	187	27	)	)	PUNCT
ajis02-348	187	28	.	.	PUNCT
ajis02-348	188	1	using	use	VERB
ajis02-348	188	2	least	least	ADJ
ajis02-348	188	3	sum	sum	NOUN
ajis02-348	188	4	of	of	ADP
ajis02-348	188	5	relative	relative	ADJ
ajis02-348	188	6	errors	error	NOUN
ajis02-348	188	7	as	as	SCONJ
ajis02-348	188	8	the	the	DET
ajis02-348	188	9	regression	regression	NOUN
ajis02-348	188	10	criterion	criterion	NOUN
ajis02-348	188	11	would	would	AUX
ajis02-348	188	12	make	make	VERB
ajis02-348	188	13	the	the	DET
ajis02-348	188	14	situation	situation	NOUN
ajis02-348	188	15	even	even	ADV
ajis02-348	188	16	more	more	ADV
ajis02-348	188	17	serious	serious	ADJ
ajis02-348	188	18	.	.	PUNCT
ajis02-348	189	1	a	a	DET
ajis02-348	189	2	new	new	ADJ
ajis02-348	189	3	regression	regression	NOUN
ajis02-348	189	4	criterion	criterion	NOUN
ajis02-348	189	5	from	from	ADP
ajis02-348	189	6	the	the	DET
ajis02-348	189	7	above	above	ADJ
ajis02-348	189	8	discussion	discussion	NOUN
ajis02-348	189	9	we	we	PRON
ajis02-348	189	10	can	can	AUX
ajis02-348	189	11	argue	argue	VERB
ajis02-348	189	12	that	that	SCONJ
ajis02-348	189	13	there	there	PRON
ajis02-348	189	14	are	be	VERB
ajis02-348	189	15	two	two	NUM
ajis02-348	189	16	special	special	ADJ
ajis02-348	189	17	requirements	requirement	NOUN
ajis02-348	189	18	for	for	ADP
ajis02-348	189	19	regression	regression	NOUN
ajis02-348	189	20	criterion	criterion	NOUN
ajis02-348	189	21	in	in	ADP
ajis02-348	189	22	formulating	formulate	VERB
ajis02-348	189	23	software	software	NOUN
ajis02-348	189	24	costing	cost	VERB
ajis02-348	189	25	models	model	NOUN
ajis02-348	189	26	.	.	PUNCT
ajis02-348	190	1	•	•	INTJ
ajis02-348	190	2	it	it	PRON
ajis02-348	190	3	should	should	AUX
ajis02-348	190	4	be	be	AUX
ajis02-348	190	5	in	in	ADP
ajis02-348	190	6	the	the	DET
ajis02-348	190	7	form	form	NOUN
ajis02-348	190	8	of	of	ADP
ajis02-348	190	9	the	the	DET
ajis02-348	190	10	ratio	ratio	NOUN
ajis02-348	190	11	of	of	ADP
ajis02-348	190	12	est	est	X
ajis02-348	190	13	{	{	PUNCT
ajis02-348	190	14	to	to	ADP
ajis02-348	190	15	act.t	act.t	PROPN
ajis02-348	190	16	.	.	PUNCT
ajis02-348	191	1	•	•	INTJ
ajis02-348	191	2	it	it	PRON
ajis02-348	191	3	should	should	AUX
ajis02-348	191	4	be	be	AUX
ajis02-348	191	5	an	an	DET
ajis02-348	191	6	unbiased	unbiased	ADJ
ajis02-348	191	7	one	one	NUM
ajis02-348	191	8	,	,	PUNCT
ajis02-348	191	9	which	which	PRON
ajis02-348	191	10	results	result	VERB
ajis02-348	191	11	in	in	ADP
ajis02-348	191	12	neither	neither	CCONJ
ajis02-348	191	13	systematic	systematic	ADJ
ajis02-348	191	14	underestimation	underestimation	NOUN
ajis02-348	191	15	nor	nor	CCONJ
ajis02-348	191	16	systematic	systematic	ADJ
ajis02-348	191	17	overestimation	overestimation	NOUN
ajis02-348	191	18	.	.	PUNCT
ajis02-348	192	1	to	to	PART
ajis02-348	192	2	meet	meet	VERB
ajis02-348	192	3	these	these	DET
ajis02-348	192	4	two	two	NUM
ajis02-348	192	5	special	special	ADJ
ajis02-348	192	6	demands	demand	NOUN
ajis02-348	192	7	,	,	PUNCT
ajis02-348	192	8	a	a	DET
ajis02-348	192	9	new	new	ADJ
ajis02-348	192	10	regression	regression	NOUN
ajis02-348	192	11	criterion	criterion	NOUN
ajis02-348	192	12	in	in	ADP
ajis02-348	192	13	formulating	formulate	VERB
ajis02-348	192	14	software	software	NOUN
ajis02-348	192	15	costing	costing	NOUN
ajis02-348	192	16	models	model	NOUN
ajis02-348	192	17	is	be	AUX
ajis02-348	192	18	proposed	propose	VERB
ajis02-348	192	19	:	:	PUNCT
ajis02-348	192	20	est	est	X
ajis02-348	192	21	it	it	PRON
ajis02-348	192	22	is	be	AUX
ajis02-348	192	23	obvious	obvious	ADJ
ajis02-348	192	24	that	that	SCONJ
ajis02-348	192	25	this	this	DET
ajis02-348	192	26	criterion	criterion	NOUN
ajis02-348	192	27	takes	take	VERB
ajis02-348	192	28	the	the	DET
ajis02-348	192	29	ratio	ratio	NOUN
ajis02-348	192	30	form	form	NOUN
ajis02-348	192	31	of	of	ADP
ajis02-348	192	32	estf	estf	NOUN
ajis02-348	192	33	to	to	PART
ajis02-348	192	34	actf	actf	ADJ
ajis02-348	192	35	.	.	PUNCT
ajis02-348	193	1	the	the	PRON
ajis02-348	193	2	closer	close	ADJ
ajis02-348	193	3	(	(	PUNCT
ajis02-348	193	4	in	in	ADP
ajis02-348	193	5	'	'	NOUN
ajis02-348	193	6	-	-	PUNCT
ajis02-348	193	7	)	)	PUNCT
ajis02-348	193	8	is	be	AUX
ajis02-348	193	9	to	to	ADP
ajis02-348	193	10	0	0	NUM
ajis02-348	193	11	,	,	PUNCT
ajis02-348	193	12	the	the	DET
ajis02-348	193	13	closer	close	ADJ
ajis02-348	193	14	act	act	NOUN
ajis02-348	193	15	(	(	PUNCT
ajis02-348	193	16	estf	estf	NOUN
ajis02-348	193	17	is	be	AUX
ajis02-348	193	18	to	to	ADP
ajis02-348	193	19	1	1	NUM
ajis02-348	193	20	,	,	PUNCT
ajis02-348	193	21	and	and	CCONJ
ajis02-348	193	22	the	the	DET
ajis02-348	193	23	more	more	ADV
ajis02-348	193	24	accurate	accurate	ADJ
ajis02-348	193	25	the	the	DET
ajis02-348	193	26	estimation	estimation	NOUN
ajis02-348	193	27	.	.	PUNCT
ajis02-348	194	1	as	as	SCONJ
ajis02-348	194	2	shown	show	VERB
ajis02-348	194	3	in	in	ADP
ajis02-348	194	4	the	the	DET
ajis02-348	194	5	above	above	ADJ
ajis02-348	194	6	expression	expression	NOUN
ajis02-348	194	7	,	,	PUNCT
ajis02-348	194	8	est	est	X
ajis02-348	194	9	;	;	PUNCT
ajis02-348	194	10	and	and	CCONJ
ajis02-348	194	11	act	act	PROPN
ajis02-348	194	12	-	-	PUNCT
ajis02-348	194	13	t	t	PROPN
ajis02-348	194	14	are	be	AUX
ajis02-348	194	15	act	act	NOUN
ajis02-348	194	16	,	,	PUNCT
ajis02-348	194	17	symmetric	symmetric	ADJ
ajis02-348	194	18	with	with	ADP
ajis02-348	194	19	respect	respect	NOUN
ajis02-348	194	20	to	to	ADP
ajis02-348	194	21	their	their	PRON
ajis02-348	194	22	positions	position	NOUN
ajis02-348	194	23	in	in	ADP
ajis02-348	194	24	the	the	DET
ajis02-348	194	25	expression	expression	NOUN
ajis02-348	194	26	.	.	PUNCT
ajis02-348	195	1	therefore	therefore	ADV
ajis02-348	195	2	,	,	PUNCT
ajis02-348	195	3	this	this	DET
ajis02-348	195	4	criterion	criterion	NOUN
ajis02-348	195	5	does	do	AUX
ajis02-348	195	6	not	not	PART
ajis02-348	195	7	produce	produce	VERB
ajis02-348	195	8	an	an	DET
ajis02-348	195	9	estimation	estimation	NOUN
ajis02-348	195	10	formula	formula	NOUN
ajis02-348	195	11	of	of	ADP
ajis02-348	195	12	either	either	CCONJ
ajis02-348	195	13	systematic	systematic	ADJ
ajis02-348	195	14	underestimation	underestimation	NOUN
ajis02-348	195	15	or	or	CCONJ
ajis02-348	195	16	systematic	systematic	ADJ
ajis02-348	195	17	overestimation	overestimation	NOUN
ajis02-348	195	18	.	.	PUNCT
ajis02-348	196	1	computational	computational	ADJ
ajis02-348	196	2	procedure	procedure	NOUN
ajis02-348	196	3	with	with	ADP
ajis02-348	196	4	the	the	DET
ajis02-348	196	5	new	new	ADJ
ajis02-348	196	6	regression	regression	NOUN
ajis02-348	196	7	criterion	criterion	NOUN
ajis02-348	196	8	we	we	PRON
ajis02-348	196	9	can	can	AUX
ajis02-348	196	10	make	make	VERB
ajis02-348	196	11	use	use	NOUN
ajis02-348	196	12	of	of	ADP
ajis02-348	196	13	existing	exist	VERB
ajis02-348	196	14	statistics	statistic	NOUN
ajis02-348	196	15	software	software	NOUN
ajis02-348	196	16	in	in	ADP
ajis02-348	196	17	performing	perform	VERB
ajis02-348	196	18	non	non	ADJ
ajis02-348	196	19	-	-	ADJ
ajis02-348	196	20	linear	linear	ADJ
ajis02-348	196	21	regression	regression	NOUN
ajis02-348	196	22	with	with	ADP
ajis02-348	196	23	the	the	DET
ajis02-348	196	24	proposed	propose	VERB
ajis02-348	196	25	criterion	criterion	NOUN
ajis02-348	196	26	.	.	PUNCT
ajis02-348	197	1	in	in	ADP
ajis02-348	197	2	statistics	statistic	NOUN
ajis02-348	197	3	software	software	PROPN
ajis02-348	197	4	,	,	PUNCT
ajis02-348	197	5	the	the	DET
ajis02-348	197	6	criterion	criterion	NOUN
ajis02-348	197	7	for	for	ADP
ajis02-348	197	8	finding	find	VERB
ajis02-348	197	9	the	the	DET
ajis02-348	197	10	optimal	optimal	ADJ
ajis02-348	197	11	parameters	parameter	NOUN
ajis02-348	197	12	is	be	AUX
ajis02-348	197	13	ls	ls	ADJ
ajis02-348	197	14	min	min	NOUN
ajis02-348	197	15	[	[	PUNCT
ajis02-348	197	16	before	before	SCONJ
ajis02-348	197	17	the	the	DET
ajis02-348	197	18	non	non	ADJ
ajis02-348	197	19	-	-	ADJ
ajis02-348	197	20	linear	linear	ADJ
ajis02-348	197	21	regression	regression	NOUN
ajis02-348	197	22	is	be	AUX
ajis02-348	197	23	undertaken	undertake	VERB
ajis02-348	197	24	,	,	PUNCT
ajis02-348	197	25	the	the	DET
ajis02-348	197	26	observed	observed	ADJ
ajis02-348	197	27	values	value	NOUN
ajis02-348	197	28	and	and	CCONJ
ajis02-348	197	29	the	the	DET
ajis02-348	197	30	functional	functional	ADJ
ajis02-348	197	31	form	form	NOUN
ajis02-348	197	32	of	of	ADP
ajis02-348	197	33	the	the	DET
ajis02-348	197	34	model	model	NOUN
ajis02-348	197	35	can	can	AUX
ajis02-348	197	36	be	be	AUX
ajis02-348	197	37	transferred	transfer	VERB
ajis02-348	197	38	into	into	ADP
ajis02-348	197	39	their	their	PRON
ajis02-348	197	40	logarithmic	logarithmic	ADJ
ajis02-348	197	41	counterparts	counterpart	NOUN
ajis02-348	197	42	,	,	PUNCT
ajis02-348	197	43	actf	actf	ADJ
ajis02-348	197	44	and	and	CCONJ
ajis02-348	197	45	est	est	X
ajis02-348	197	46	{	{	PUNCT
ajis02-348	197	47	.the	.the	PUNCT
ajis02-348	198	1	ls	ls	PROPN
ajis02-348	198	2	can	can	AUX
ajis02-348	198	3	be	be	AUX
ajis02-348	198	4	expressed	express	VERB
ajis02-348	198	5	as	as	ADP
ajis02-348	198	6	n	n	ADP
ajis02-348	198	7	n	n	CCONJ
ajis02-348	198	8	n	n	PRON
ajis02-348	198	9	art	art	NOUN
ajis02-348	198	10	v	v	NOUN
ajis02-348	198	11	-	-	PUNCT
ajis02-348	198	12	i	i	PRON
ajis02-348	198	13	•	•	NOUN
ajis02-348	198	14	1	1	NUM
ajis02-348	198	15	7	7	NUM
ajis02-348	198	16	v"1	v"1	PROPN
ajis02-348	198	17	n	n	PRON
ajis02-348	198	18	\	\	NOUN
ajis02-348	198	19	«	«	PUNCT
ajis02-348	198	20	\-i2	\-i2	X
ajis02-348	198	21	x^	x^	VERB
ajis02-348	198	22	11	11	NUM
ajis02-348	198	23	eal	eal	NOUN
ajis02-348	198	24	:	:	PUNCT
ajis02-348	198	25	2	2	NUM
ajis02-348	198	26	min	min	NOUN
ajis02-348	198	27	[	[	PUNCT
ajis02-348	198	28	>	>	X
ajis02-348	198	29	(	(	PUNCT
ajis02-348	198	30	actf	actf	ADJ
ajis02-348	198	31	estf	estf	NOUN
ajis02-348	198	32	)	)	PUNCT
ajis02-348	198	33	]	]	PUNCT
ajis02-348	199	1	=	=	PUNCT
ajis02-348	199	2	min	min	NOUN
ajis02-348	199	3	{	{	PUNCT
ajis02-348	199	4	>	>	X
ajis02-348	200	1	[	[	X
ajis02-348	200	2	m(act	m(act	PROPN
ajis02-348	200	3	.	.	PUNCT
ajis02-348	200	4	)	)	PUNCT
ajis02-348	200	5	m(esti	m(esti	PROPN
ajis02-348	200	6	.	.	PUNCT
ajis02-348	200	7	)	)	PUNCT
ajis02-348	200	8	]	]	PUNCT
ajis02-348	200	9	}	}	PUNCT
ajis02-348	200	10	=	=	SYM
ajis02-348	200	11	min	min	NOUN
ajis02-348	200	12	[	[	PUNCT
ajis02-348	200	13	>	>	X
ajis02-348	200	14	(	(	PUNCT
ajis02-348	200	15	in	in	ADP
ajis02-348	200	16	-	-	PUNCT
ajis02-348	200	17	)	)	PUNCT
ajis02-348	200	18	]	]	PUNCT
ajis02-348	200	19	for	for	ADP
ajis02-348	200	20	example	example	NOUN
ajis02-348	200	21	,	,	PUNCT
ajis02-348	200	22	we	we	PRON
ajis02-348	200	23	have	have	VERB
ajis02-348	200	24	a	a	DET
ajis02-348	200	25	set	set	NOUN
ajis02-348	200	26	of	of	ADP
ajis02-348	200	27	n	n	DET
ajis02-348	200	28	pairs	pair	NOUN
ajis02-348	200	29	of	of	ADP
ajis02-348	200	30	data	datum	NOUN
ajis02-348	200	31	(	(	PUNCT
ajis02-348	200	32	size	size	NOUN
ajis02-348	200	33	t	t	PROPN
ajis02-348	200	34	,	,	PUNCT
ajis02-348	200	35	effort	effort	NOUN
ajis02-348	200	36	.	.	PUNCT
ajis02-348	200	37	)	)	PUNCT
ajis02-348	201	1	(	(	PUNCT
ajis02-348	201	2	i	i	NOUN
ajis02-348	201	3	=	=	PUNCT
ajis02-348	201	4	1,2,3,	1,2,3,	NUM
ajis02-348	201	5	...	...	PUNCT
ajis02-348	201	6	,n	,n	NOUN
ajis02-348	201	7	)	)	PUNCT
ajis02-348	201	8	.	.	PUNCT
ajis02-348	202	1	if	if	SCONJ
ajis02-348	202	2	the	the	DET
ajis02-348	202	3	functional	functional	ADJ
ajis02-348	202	4	form	form	NOUN
ajis02-348	202	5	of	of	ADP
ajis02-348	202	6	a	a	DET
ajis02-348	202	7	model	model	NOUN
ajis02-348	202	8	is	be	AUX
ajis02-348	202	9	effort	effort	NOUN
ajis02-348	202	10	=	=	SYM
ajis02-348	202	11	a(size	a(size	VERB
ajis02-348	202	12	)	)	PUNCT
ajis02-348	203	1	+	+	SYM
ajis02-348	203	2	b	b	NOUN
ajis02-348	203	3	ajis	ajis	ADV
ajis02-348	203	4	vol	vol	VERB
ajis02-348	203	5	5	5	NUM
ajis02-348	203	6	no	no	DET
ajis02-348	203	7	1	1	NUM
ajis02-348	203	8	september	september	NOUN
ajis02-348	203	9	1997	1997	NUM
ajis02-348	203	10	38	38	NUM
ajis02-348	203	11	we	we	PRON
ajis02-348	203	12	can	can	AUX
ajis02-348	203	13	take	take	VERB
ajis02-348	203	14	the	the	DET
ajis02-348	203	15	logarithmic	logarithmic	ADJ
ajis02-348	203	16	form	form	NOUN
ajis02-348	203	17	for	for	ADP
ajis02-348	203	18	the	the	DET
ajis02-348	203	19	both	both	DET
ajis02-348	203	20	sides	side	NOUN
ajis02-348	203	21	of	of	ADP
ajis02-348	203	22	the	the	DET
ajis02-348	203	23	equation	equation	NOUN
ajis02-348	203	24	to	to	PART
ajis02-348	203	25	obtain	obtain	VERB
ajis02-348	203	26	in(effort	in(effort	NOUN
ajis02-348	203	27	)	)	PUNCT
ajis02-348	203	28	=	=	SYM
ajis02-348	203	29	\n[a(size	\n[a(size	ADJ
ajis02-348	203	30	)	)	PUNCT
ajis02-348	204	1	+	+	SYM
ajis02-348	205	1	b	b	X
ajis02-348	205	2	]	]	X
ajis02-348	205	3	.	.	PUNCT
ajis02-348	206	1	the	the	DET
ajis02-348	206	2	regression	regression	NOUN
ajis02-348	206	3	is	be	AUX
ajis02-348	206	4	to	to	PART
ajis02-348	206	5	be	be	AUX
ajis02-348	206	6	performed	perform	VERB
ajis02-348	206	7	according	accord	VERB
ajis02-348	206	8	to	to	ADP
ajis02-348	206	9	this	this	DET
ajis02-348	206	10	equation	equation	NOUN
ajis02-348	206	11	.	.	PUNCT
ajis02-348	207	1	in	in	ADP
ajis02-348	207	2	this	this	DET
ajis02-348	207	3	case	case	NOUN
ajis02-348	207	4	,	,	PUNCT
ajis02-348	207	5	the	the	DET
ajis02-348	207	6	/	/	SYM
ajis02-348	207	7	'	'	NOUN
ajis02-348	207	8	*	*	PUNCT
ajis02-348	207	9	estimate	estimate	NOUN
ajis02-348	207	10	is	be	AUX
ajis02-348	207	11	esti	esti	PROPN
ajis02-348	207	12	=	=	PUNCT
ajis02-348	207	13	a(sizej	a(sizej	NOUN
ajis02-348	207	14	)	)	PUNCT
ajis02-348	208	1	+	+	CCONJ
ajis02-348	208	2	b	b	NOUN
ajis02-348	208	3	and	and	CCONJ
ajis02-348	208	4	the	the	DET
ajis02-348	208	5	actual	actual	ADJ
ajis02-348	208	6	value	value	NOUN
ajis02-348	208	7	is	be	AUX
ajis02-348	208	8	acti	acti	VERB
ajis02-348	208	9	=	=	SYM
ajis02-348	208	10	effort	effort	NOUN
ajis02-348	208	11	,	,	PUNCT
ajis02-348	208	12	.	.	PUNCT
ajis02-348	209	1	before	before	ADP
ajis02-348	209	2	the	the	DET
ajis02-348	209	3	regression	regression	NOUN
ajis02-348	209	4	,	,	PUNCT
ajis02-348	209	5	we	we	PRON
ajis02-348	209	6	first	first	ADV
ajis02-348	209	7	calculate	calculate	VERB
ajis02-348	209	8	ln(effort	ln(effort	PROPN
ajis02-348	209	9	:	:	PUNCT
ajis02-348	209	10	)	)	PUNCT
ajis02-348	209	11	(	(	PUNCT
ajis02-348	209	12	i=	i=	PROPN
ajis02-348	209	13	l,2,3,	l,2,3,	NUM
ajis02-348	209	14	...	...	PUNCT
ajis02-348	209	15	,n	,n	NUM
ajis02-348	209	16	)	)	PUNCT
ajis02-348	209	17	,	,	PUNCT
ajis02-348	209	18	which	which	PRON
ajis02-348	209	19	will	will	AUX
ajis02-348	209	20	be	be	AUX
ajis02-348	209	21	used	use	VERB
ajis02-348	209	22	as	as	ADP
ajis02-348	209	23	the	the	DET
ajis02-348	209	24	values	value	NOUN
ajis02-348	209	25	for	for	ADP
ajis02-348	209	26	the	the	DET
ajis02-348	209	27	dependent	dependent	ADJ
ajis02-348	209	28	variable	variable	NOUN
ajis02-348	209	29	for	for	ADP
ajis02-348	209	30	the	the	DET
ajis02-348	209	31	regression	regression	NOUN
ajis02-348	209	32	.	.	PUNCT
ajis02-348	210	1	in	in	ADP
ajis02-348	210	2	the	the	DET
ajis02-348	210	3	regression	regression	NOUN
ajis02-348	210	4	,	,	PUNCT
ajis02-348	210	5	the	the	DET
ajis02-348	210	6	functional	functional	ADJ
ajis02-348	210	7	form	form	NOUN
ajis02-348	210	8	is	be	AUX
ajis02-348	210	9	ln[fl(s	ln[fl(s	PROPN
ajis02-348	210	10	/	/	SYM
ajis02-348	210	11	ze	ze	PROPN
ajis02-348	210	12	)	)	PUNCT
ajis02-348	210	13	4b	4b	X
ajis02-348	210	14	]	]	PUNCT
ajis02-348	210	15	.	.	PUNCT
ajis02-348	211	1	therefore	therefore	ADV
ajis02-348	211	2	we	we	PRON
ajis02-348	211	3	have	have	VERB
ajis02-348	211	4	the	the	DET
ajis02-348	211	5	criterion	criterion	NOUN
ajis02-348	211	6	mm	mm	PROPN
ajis02-348	211	7	i2	i2	PROPN
ajis02-348	211	8	.	.	PUNCT
ajis02-348	211	9	]	]	PUNCT
ajis02-348	211	10	}	}	PUNCT
ajis02-348	211	11	effort	effort	NOUN
ajis02-348	211	12	,	,	PUNCT
ajis02-348	211	13	assessment	assessment	NOUN
ajis02-348	211	14	of	of	ADP
ajis02-348	211	15	the	the	DET
ajis02-348	211	16	proposed	propose	VERB
ajis02-348	211	17	measures	measure	NOUN
ajis02-348	211	18	and	and	CCONJ
ajis02-348	211	19	criterion	criterion	NOUN
ajis02-348	211	20	a	a	DET
ajis02-348	211	21	set	set	NOUN
ajis02-348	211	22	of	of	ADP
ajis02-348	211	23	data	datum	NOUN
ajis02-348	211	24	quoted	quote	VERB
ajis02-348	211	25	by	by	ADP
ajis02-348	211	26	desharnais	desharnais	NOUN
ajis02-348	211	27	(	(	PUNCT
ajis02-348	211	28	1988	1988	NUM
ajis02-348	211	29	)	)	PUNCT
ajis02-348	211	30	from	from	ADP
ajis02-348	211	31	81	81	NUM
ajis02-348	211	32	software	software	NOUN
ajis02-348	211	33	projects	project	NOUN
ajis02-348	211	34	was	be	AUX
ajis02-348	211	35	used	use	VERB
ajis02-348	211	36	to	to	PART
ajis02-348	211	37	assess	assess	VERB
ajis02-348	211	38	these	these	DET
ajis02-348	211	39	measures	measure	NOUN
ajis02-348	211	40	of	of	ADP
ajis02-348	211	41	estimation	estimation	NOUN
ajis02-348	211	42	accuracy	accuracy	NOUN
ajis02-348	211	43	and	and	CCONJ
ajis02-348	211	44	consistency	consistency	NOUN
ajis02-348	211	45	,	,	PUNCT
ajis02-348	211	46	and	and	CCONJ
ajis02-348	211	47	regression	regression	NOUN
ajis02-348	211	48	criteria	criterion	NOUN
ajis02-348	211	49	.	.	PUNCT
ajis02-348	212	1	the	the	DET
ajis02-348	212	2	main	main	ADJ
ajis02-348	212	3	features	feature	NOUN
ajis02-348	212	4	of	of	ADP
ajis02-348	212	5	the	the	DET
ajis02-348	212	6	data	datum	NOUN
ajis02-348	212	7	set	set	VERB
ajis02-348	212	8	are	be	AUX
ajis02-348	212	9	summarised	summarise	VERB
ajis02-348	212	10	below	below	ADV
ajis02-348	212	11	.	.	PUNCT
ajis02-348	213	1	effort	effort	NOUN
ajis02-348	213	2	range	range	NOUN
ajis02-348	213	3	:	:	PUNCT
ajis02-348	213	4	546	546	NUM
ajis02-348	213	5	-	-	SYM
ajis02-348	213	6	19,894	19,894	NUM
ajis02-348	213	7	person	person	NOUN
ajis02-348	213	8	-	-	PUNCT
ajis02-348	213	9	hours	hour	NOUN
ajis02-348	213	10	function	function	NOUN
ajis02-348	213	11	point	point	NOUN
ajis02-348	213	12	range	range	NOUN
ajis02-348	213	13	:	:	PUNCT
ajis02-348	213	14	621	621	NUM
ajis02-348	213	15	,	,	PUNCT
ajis02-348	213	16	1	1	NUM
ajis02-348	213	17	1	1	NUM
ajis02-348	213	18	6	6	NUM
ajis02-348	213	19	fp	fp	PROPN
ajis02-348	213	20	team	team	NOUN
ajis02-348	213	21	average	average	ADJ
ajis02-348	213	22	experience	experience	NOUN
ajis02-348	213	23	:	:	PUNCT
ajis02-348	213	24	0	0	NUM
ajis02-348	213	25	-	-	SYM
ajis02-348	213	26	4	4	NUM
ajis02-348	213	27	years	year	NOUN
ajis02-348	213	28	manager	manager	NOUN
ajis02-348	213	29	experience	experience	NOUN
ajis02-348	213	30	:	:	PUNCT
ajis02-348	213	31	0	0	NUM
ajis02-348	213	32	-	-	SYM
ajis02-348	213	33	7	7	NUM
ajis02-348	213	34	years	year	NOUN
ajis02-348	213	35	language	language	NOUN
ajis02-348	213	36	levels	level	NOUN
ajis02-348	213	37	:	:	PUNCT
ajis02-348	213	38	level	level	NOUN
ajis02-348	213	39	1	1	NUM
ajis02-348	213	40	:	:	SYM
ajis02-348	213	41	46	46	NUM
ajis02-348	213	42	projects	project	NOUN
ajis02-348	213	43	were	be	AUX
ajis02-348	213	44	developed	develop	VERB
ajis02-348	213	45	in	in	ADP
ajis02-348	213	46	3gl	3gl	NOUN
ajis02-348	213	47	(	(	PUNCT
ajis02-348	213	48	cobol	cobol	NOUN
ajis02-348	213	49	with	with	ADP
ajis02-348	213	50	ims	im	NOUN
ajis02-348	213	51	or	or	CCONJ
ajis02-348	213	52	idms	idms	ADJ
ajis02-348	213	53	type	type	NOUN
ajis02-348	213	54	databases	database	NOUN
ajis02-348	213	55	)	)	PUNCT
ajis02-348	213	56	.	.	PUNCT
ajis02-348	214	1	level	level	NOUN
ajis02-348	214	2	2	2	NUM
ajis02-348	214	3	:	:	SYM
ajis02-348	214	4	25	25	NUM
ajis02-348	214	5	projects	project	NOUN
ajis02-348	214	6	were	be	AUX
ajis02-348	214	7	developed	develop	VERB
ajis02-348	214	8	using	use	VERB
ajis02-348	214	9	a	a	DET
ajis02-348	214	10	combination	combination	NOUN
ajis02-348	214	11	of	of	ADP
ajis02-348	214	12	the	the	DET
ajis02-348	214	13	traditional	traditional	ADJ
ajis02-348	214	14	3gl	3gl	ADJ
ajis02-348	214	15	approach	approach	NOUN
ajis02-348	214	16	together	together	ADV
ajis02-348	214	17	with	with	ADP
ajis02-348	214	18	screen	screen	NOUN
ajis02-348	214	19	and	and	CCONJ
ajis02-348	214	20	report	report	NOUN
ajis02-348	214	21	generators	generator	NOUN
ajis02-348	214	22	.	.	PUNCT
ajis02-348	215	1	level	level	NOUN
ajis02-348	215	2	3	3	NUM
ajis02-348	215	3	:	:	SYM
ajis02-348	215	4	10	10	NUM
ajis02-348	215	5	projects	project	NOUN
ajis02-348	215	6	were	be	AUX
ajis02-348	215	7	implemented	implement	VERB
ajis02-348	215	8	in	in	ADP
ajis02-348	215	9	4gl	4gl	PROPN
ajis02-348	215	10	.	.	PUNCT
ajis02-348	216	1	integrated	integrate	VERB
ajis02-348	216	2	software	software	NOUN
ajis02-348	216	3	cost	cost	NOUN
ajis02-348	216	4	model	model	NOUN
ajis02-348	216	5	an	an	DET
ajis02-348	216	6	integrated	integrate	VERB
ajis02-348	216	7	software	software	NOUN
ajis02-348	216	8	cost	cost	NOUN
ajis02-348	216	9	model	model	NOUN
ajis02-348	216	10	was	be	AUX
ajis02-348	216	11	proposed	propose	VERB
ajis02-348	216	12	by	by	ADP
ajis02-348	216	13	gao	gao	PROPN
ajis02-348	216	14	and	and	CCONJ
ajis02-348	216	15	lo	lo	PROPN
ajis02-348	216	16	(	(	PUNCT
ajis02-348	216	17	1995	1995	NUM
ajis02-348	216	18	)	)	PUNCT
ajis02-348	216	19	on	on	ADP
ajis02-348	216	20	the	the	DET
ajis02-348	216	21	basis	basis	NOUN
ajis02-348	216	22	of	of	ADP
ajis02-348	216	23	cocomo	cocomo	NOUN
ajis02-348	216	24	(	(	PUNCT
ajis02-348	216	25	boehm	boehm	NOUN
ajis02-348	216	26	,	,	PUNCT
ajis02-348	216	27	1981	1981	NUM
ajis02-348	216	28	)	)	PUNCT
ajis02-348	216	29	and	and	CCONJ
ajis02-348	216	30	fpa	fpa	PROPN
ajis02-348	216	31	.	.	PUNCT
ajis02-348	217	1	the	the	DET
ajis02-348	217	2	integrated	integrate	VERB
ajis02-348	217	3	model	model	NOUN
ajis02-348	217	4	attempts	attempt	VERB
ajis02-348	217	5	to	to	PART
ajis02-348	217	6	combine	combine	VERB
ajis02-348	217	7	the	the	DET
ajis02-348	217	8	advantages	advantage	NOUN
ajis02-348	217	9	of	of	ADP
ajis02-348	217	10	these	these	DET
ajis02-348	217	11	two	two	NUM
ajis02-348	217	12	approaches	approach	NOUN
ajis02-348	217	13	into	into	ADP
ajis02-348	217	14	a	a	DET
ajis02-348	217	15	single	single	ADJ
ajis02-348	217	16	model	model	NOUN
ajis02-348	217	17	.	.	PUNCT
ajis02-348	218	1	to	to	PART
ajis02-348	218	2	address	address	VERB
ajis02-348	218	3	the	the	DET
ajis02-348	218	4	problem	problem	NOUN
ajis02-348	218	5	of	of	ADP
ajis02-348	218	6	language	language	NOUN
ajis02-348	218	7	dependency	dependency	NOUN
ajis02-348	218	8	,	,	PUNCT
ajis02-348	218	9	language	language	NOUN
ajis02-348	218	10	-	-	PUNCT
ajis02-348	218	11	weighted	weight	VERB
ajis02-348	218	12	function	function	NOUN
ajis02-348	218	13	point	point	NOUN
ajis02-348	218	14	is	be	AUX
ajis02-348	218	15	introduced	introduce	VERB
ajis02-348	218	16	.	.	PUNCT
ajis02-348	219	1	in	in	ADP
ajis02-348	219	2	addition	addition	NOUN
ajis02-348	219	3	,	,	PUNCT
ajis02-348	219	4	continuous	continuous	ADJ
ajis02-348	219	5	,	,	PUNCT
ajis02-348	219	6	rather	rather	ADV
ajis02-348	219	7	than	than	ADP
ajis02-348	219	8	discrete	discrete	ADJ
ajis02-348	219	9	,	,	PUNCT
ajis02-348	219	10	"	"	PUNCT
ajis02-348	219	11	cost	cost	NOUN
ajis02-348	219	12	drivers	driver	NOUN
ajis02-348	219	13	"	"	PUNCT
ajis02-348	219	14	are	be	AUX
ajis02-348	219	15	used	use	VERB
ajis02-348	219	16	.	.	PUNCT
ajis02-348	220	1	the	the	DET
ajis02-348	220	2	general	general	ADJ
ajis02-348	220	3	form	form	NOUN
ajis02-348	220	4	of	of	ADP
ajis02-348	220	5	the	the	DET
ajis02-348	220	6	model	model	NOUN
ajis02-348	220	7	is	be	AUX
ajis02-348	220	8	effort	effort	NOUN
ajis02-348	220	9	=	=	SYM
ajis02-348	220	10	a(wf/))*fjc	a(wf/))*fjc	ADJ
ajis02-348	220	11	,	,	PUNCT
ajis02-348	220	12	u'"'/i	u'"'/i	PROPN
ajis02-348	220	13	)	)	PUNCT
ajis02-348	220	14	where	where	SCONJ
ajis02-348	220	15	a	a	PRON
ajis02-348	220	16	and	and	CCONJ
ajis02-348	220	17	b	b	NOUN
ajis02-348	220	18	ant	ant	ADJ
ajis02-348	220	19	constants	constant	NOUN
ajis02-348	220	20	,	,	PUNCT
ajis02-348	220	21	xi	xi	X
ajis02-348	220	22	is	be	AUX
ajis02-348	220	23	the	the	DET
ajis02-348	220	24	magnitude	magnitude	NOUN
ajis02-348	220	25	of	of	ADP
ajis02-348	220	26	the	the	DET
ajis02-348	220	27	i	i	PROPN
ajis02-348	220	28	'	'	PART
ajis02-348	220	29	cost	cost	NOUN
ajis02-348	220	30	driver	driver	NOUN
ajis02-348	220	31	,	,	PUNCT
ajis02-348	220	32	c	c	NOUN
ajis02-348	220	33	,	,	PUNCT
ajis02-348	220	34	and	and	CCONJ
ajis02-348	220	35	di	di	NOUN
ajis02-348	220	36	are	be	AUX
ajis02-348	220	37	constants	constant	NOUN
ajis02-348	220	38	corresponding	correspond	VERB
ajis02-348	220	39	to	to	ADP
ajis02-348	220	40	the	the	DET
ajis02-348	220	41	i	i	PROPN
ajis02-348	220	42	cost	cost	VERB
ajis02-348	220	43	driver	driver	NOUN
ajis02-348	220	44	,	,	PUNCT
ajis02-348	220	45	n	n	PRON
ajis02-348	220	46	is	be	AUX
ajis02-348	220	47	the	the	DET
ajis02-348	220	48	number	number	NOUN
ajis02-348	220	49	of	of	ADP
ajis02-348	220	50	cost	cost	NOUN
ajis02-348	220	51	drivers	driver	NOUN
ajis02-348	220	52	and	and	CCONJ
ajis02-348	220	53	wfp	wfp	PROPN
ajis02-348	220	54	is	be	AUX
ajis02-348	220	55	language	language	NOUN
ajis02-348	220	56	-	-	PUNCT
ajis02-348	220	57	weighted	weight	VERB
ajis02-348	220	58	fp	fp	NOUN
ajis02-348	220	59	.	.	PUNCT
ajis02-348	221	1	the	the	DET
ajis02-348	221	2	equation	equation	NOUN
ajis02-348	221	3	appropriate	appropriate	ADJ
ajis02-348	221	4	to	to	ADP
ajis02-348	221	5	desharnairs	desharnair	NOUN
ajis02-348	221	6	'	'	PART
ajis02-348	221	7	data	datum	NOUN
ajis02-348	221	8	set	set	VERB
ajis02-348	221	9	is	be	AUX
ajis02-348	221	10	effort	effort	NOUN
ajis02-348	221	11	=	=	SYM
ajis02-348	221	12	a(lkfp)b	a(lkfp)b	PROPN
ajis02-348	221	13	.	.	PUNCT
ajis02-348	222	1	where	where	SCONJ
ajis02-348	222	2	/	/	SYM
ajis02-348	222	3	,	,	PUNCT
ajis02-348	222	4	=	=	NOUN
ajis02-348	222	5	1.00	1.00	NUM
ajis02-348	222	6	,	,	PUNCT
ajis02-348	222	7	l^	l^	NOUN
ajis02-348	222	8	=	=	SYM
ajis02-348	222	9	0.95	0.95	NUM
ajis02-348	222	10	and	and	CCONJ
ajis02-348	222	11	l±	l±	PUNCT
ajis02-348	222	12	=	=	NOUN
ajis02-348	222	13	0.20	0.20	NUM
ajis02-348	222	14	are	be	AUX
ajis02-348	222	15	the	the	DET
ajis02-348	222	16	weights	weight	NOUN
ajis02-348	222	17	for	for	ADP
ajis02-348	222	18	language	language	NOUN
ajis02-348	222	19	level	level	NOUN
ajis02-348	222	20	1	1	NUM
ajis02-348	222	21	,	,	PUNCT
ajis02-348	222	22	2	2	NUM
ajis02-348	222	23	,	,	PUNCT
ajis02-348	222	24	and	and	CCONJ
ajis02-348	222	25	3	3	NUM
ajis02-348	222	26	respectively	respectively	ADV
ajis02-348	222	27	(	(	PUNCT
ajis02-348	222	28	gao	gao	PROPN
ajis02-348	222	29	et	et	PROPN
ajis02-348	222	30	al	al	PROPN
ajis02-348	222	31	,	,	PUNCT
ajis02-348	222	32	1995	1995	NUM
ajis02-348	222	33	)	)	PUNCT
ajis02-348	222	34	.	.	PUNCT
ajis02-348	223	1	the	the	DET
ajis02-348	223	2	exponent	exponent	PROPN
ajis02-348	223	3	b	b	PROPN
ajis02-348	223	4	is	be	AUX
ajis02-348	223	5	the	the	DET
ajis02-348	223	6	responsiveness	responsiveness	NOUN
ajis02-348	223	7	of	of	ADP
ajis02-348	223	8	effort	effort	NOUN
ajis02-348	223	9	to	to	ADP
ajis02-348	223	10	fp	fp	X
ajis02-348	223	11	.	.	PUNCT
ajis02-348	224	1	it	it	PRON
ajis02-348	224	2	indicates	indicate	VERB
ajis02-348	224	3	that	that	SCONJ
ajis02-348	224	4	effort	effort	NOUN
ajis02-348	224	5	increases	increase	NOUN
ajis02-348	224	6	by	by	ADP
ajis02-348	224	7	b%	b%	NOUN
ajis02-348	224	8	when	when	SCONJ
ajis02-348	224	9	fp	fp	X
ajis02-348	224	10	increases	increase	NOUN
ajis02-348	224	11	by	by	ADP
ajis02-348	224	12	1	1	NUM
ajis02-348	224	13	%	%	NOUN
ajis02-348	224	14	.	.	PUNCT
ajis02-348	225	1	ajis	ajis	ADV
ajis02-348	225	2	vol	vol	VERB
ajis02-348	225	3	5	5	NUM
ajis02-348	225	4	no	no	DET
ajis02-348	225	5	1	1	NUM
ajis02-348	225	6	september	september	NOUN
ajis02-348	225	7	1997	1997	NUM
ajis02-348	225	8	39	39	NUM
ajis02-348	225	9	assessment	assessment	NOUN
ajis02-348	225	10	on	on	ADP
ajis02-348	225	11	regression	regression	NOUN
ajis02-348	225	12	criteria	criterion	NOUN
ajis02-348	225	13	three	three	NUM
ajis02-348	225	14	criteria	criterion	NOUN
ajis02-348	225	15	for	for	ADP
ajis02-348	225	16	finding	find	VERB
ajis02-348	225	17	optimal	optimal	ADJ
ajis02-348	225	18	parameters	parameter	NOUN
ajis02-348	225	19	are	be	AUX
ajis02-348	225	20	used	use	VERB
ajis02-348	225	21	in	in	ADP
ajis02-348	225	22	regression	regression	NOUN
ajis02-348	225	23	:	:	PUNCT
ajis02-348	225	24	•	•	NUM
ajis02-348	225	25	criterion	criterion	NOUN
ajis02-348	225	26	1	1	NUM
ajis02-348	225	27	:	:	PUNCT
ajis02-348	225	28	min[^	min[^	PROPN
ajis02-348	225	29	(	(	PUNCT
ajis02-348	225	30	act	act	PROPN
ajis02-348	225	31	(	(	PUNCT
ajis02-348	225	32	—	—	PUNCT
ajis02-348	225	33	esti	esti	X
ajis02-348	225	34	)	)	PUNCT
ajis02-348	225	35	2	2	NUM
ajis02-348	225	36	]	]	PUNCT
ajis02-348	225	37	1=1	1=1	NUM
ajis02-348	225	38	^	^	SYM
ajis02-348	225	39	.act.-est,^	.act.-est,^	PROPN
ajis02-348	225	40	•	•	NUM
ajis02-348	225	41	cnterion	cnterion	NOUN
ajis02-348	225	42	2	2	NUM
ajis02-348	225	43	:	:	PUNCT
ajis02-348	225	44	min	min	PROPN
ajis02-348	225	45	[	[	PUNCT
ajis02-348	225	46	>	>	X
ajis02-348	225	47	(	(	PUNCT
ajis02-348	225	48	)	)	PUNCT
ajis02-348	225	49	]	]	PUNCT
ajis02-348	225	50	and	and	CCONJ
ajis02-348	225	51	ti	ti	X
ajis02-348	225	52	acti	acti	VERB
ajis02-348	225	53	v^	v^	NOUN
ajis02-348	225	54	,	,	PUNCT
ajis02-348	225	55	,	,	PUNCT
ajis02-348	225	56	est	est	X
ajis02-348	225	57	:	:	PUNCT
ajis02-348	225	58	.	.	PUNCT
ajis02-348	226	1	•	•	NUM
ajis02-348	226	2	criterion	criterion	NOUN
ajis02-348	226	3	3	3	NUM
ajis02-348	226	4	:	:	SYM
ajis02-348	226	5	min	min	PROPN
ajis02-348	226	6	[	[	PUNCT
ajis02-348	226	7	2^	2^	NUM
ajis02-348	226	8	(	(	PUNCT
ajis02-348	226	9	in	in	ADP
ajis02-348	226	10	)	)	PUNCT
ajis02-348	226	11	]	]	PUNCT
ajis02-348	226	12	in	in	ADP
ajis02-348	226	13	figure	figure	NOUN
ajis02-348	226	14	2	2	NUM
ajis02-348	226	15	and	and	CCONJ
ajis02-348	226	16	figure	figure	VERB
ajis02-348	226	17	3	3	NUM
ajis02-348	226	18	,	,	PUNCT
ajis02-348	226	19	the	the	DET
ajis02-348	226	20	relative	relative	ADJ
ajis02-348	226	21	errors	error	NOUN
ajis02-348	226	22	re	re	VERB
ajis02-348	226	23	,	,	PUNCT
ajis02-348	226	24	=	=	SYM
ajis02-348	226	25	•	•	NUM
ajis02-348	226	26	est	est	X
ajis02-348	226	27	{	{	PUNCT
ajis02-348	226	28	acti	acti	PROPN
ajis02-348	226	29	act	act	PROPN
ajis02-348	226	30	.	.	PUNCT
ajis02-348	227	1	(	(	PUNCT
ajis02-348	227	2	i=	i=	PROPN
ajis02-348	227	3	1.2,3	1.2,3	NUM
ajis02-348	227	4	81	81	NUM
ajis02-348	227	5	)	)	PUNCT
ajis02-348	227	6	are	be	AUX
ajis02-348	227	7	shown	show	VERB
ajis02-348	227	8	in	in	ADP
ajis02-348	227	9	an	an	DET
ajis02-348	227	10	ascending	ascend	VERB
ajis02-348	227	11	order	order	NOUN
ajis02-348	227	12	.	.	PUNCT
ajis02-348	228	1	figure	figure	NOUN
ajis02-348	228	2	2	2	NUM
ajis02-348	228	3	shows	show	VERB
ajis02-348	228	4	the	the	DET
ajis02-348	228	5	relative	relative	ADJ
ajis02-348	228	6	errors	error	NOUN
ajis02-348	228	7	with	with	ADP
ajis02-348	228	8	respect	respect	NOUN
ajis02-348	228	9	to	to	ADP
ajis02-348	228	10	criterion	criterion	NOUN
ajis02-348	228	11	2	2	NUM
ajis02-348	228	12	.	.	PUNCT
ajis02-348	229	1	it	it	PRON
ajis02-348	229	2	can	can	AUX
ajis02-348	229	3	be	be	AUX
ajis02-348	229	4	seen	see	VERB
ajis02-348	229	5	that	that	SCONJ
ajis02-348	229	6	more	more	ADJ
ajis02-348	229	7	than	than	ADP
ajis02-348	229	8	70	70	NUM
ajis02-348	229	9	%	%	NOUN
ajis02-348	229	10	of	of	ADP
ajis02-348	229	11	the	the	DET
ajis02-348	229	12	relative	relative	ADJ
ajis02-348	229	13	errors	error	NOUN
ajis02-348	229	14	are	be	AUX
ajis02-348	229	15	negative	negative	ADJ
ajis02-348	229	16	.	.	PUNCT
ajis02-348	230	1	this	this	PRON
ajis02-348	230	2	indicates	indicate	VERB
ajis02-348	230	3	that	that	SCONJ
ajis02-348	230	4	the	the	DET
ajis02-348	230	5	formula	formula	NOUN
ajis02-348	230	6	obtained	obtain	VERB
ajis02-348	230	7	from	from	ADP
ajis02-348	230	8	criterion	criterion	NOUN
ajis02-348	230	9	2	2	NUM
ajis02-348	230	10	tends	tend	VERB
ajis02-348	230	11	towards	towards	ADP
ajis02-348	230	12	underestimation	underestimation	NOUN
ajis02-348	230	13	.	.	PUNCT
ajis02-348	231	1	figure	figure	NOUN
ajis02-348	231	2	3	3	NUM
ajis02-348	231	3	shows	show	VERB
ajis02-348	231	4	the	the	DET
ajis02-348	231	5	relative	relative	ADJ
ajis02-348	231	6	errors	error	NOUN
ajis02-348	231	7	with	with	ADP
ajis02-348	231	8	respect	respect	NOUN
ajis02-348	231	9	to	to	ADP
ajis02-348	231	10	criterion	criterion	NOUN
ajis02-348	231	11	3	3	NUM
ajis02-348	231	12	.	.	PUNCT
ajis02-348	232	1	the	the	DET
ajis02-348	232	2	number	number	NOUN
ajis02-348	232	3	of	of	ADP
ajis02-348	232	4	negative	negative	ADJ
ajis02-348	232	5	relative	relative	ADJ
ajis02-348	232	6	errors	error	NOUN
ajis02-348	232	7	is	be	AUX
ajis02-348	232	8	approximately	approximately	ADV
ajis02-348	232	9	the	the	DET
ajis02-348	232	10	same	same	ADJ
ajis02-348	232	11	as	as	ADP
ajis02-348	232	12	that	that	PRON
ajis02-348	232	13	of	of	ADP
ajis02-348	232	14	positive	positive	ADJ
ajis02-348	232	15	relative	relative	ADJ
ajis02-348	232	16	errors	error	NOUN
ajis02-348	232	17	.	.	PUNCT
ajis02-348	233	1	the	the	DET
ajis02-348	233	2	formula	formula	NOUN
ajis02-348	233	3	obtained	obtain	VERB
ajis02-348	233	4	from	from	ADP
ajis02-348	233	5	criterion	criterion	NOUN
ajis02-348	233	6	3	3	NUM
ajis02-348	233	7	does	do	AUX
ajis02-348	233	8	not	not	PART
ajis02-348	233	9	result	result	VERB
ajis02-348	233	10	in	in	ADP
ajis02-348	233	11	systematic	systematic	ADJ
ajis02-348	233	12	bias	bias	NOUN
ajis02-348	233	13	.	.	PUNCT
ajis02-348	234	1	(	(	PUNCT
ajis02-348	234	2	criterion	criterion	NOUN
ajis02-348	234	3	1	1	NUM
ajis02-348	234	4	is	be	AUX
ajis02-348	234	5	similar	similar	ADJ
ajis02-348	234	6	to	to	ADP
ajis02-348	234	7	criterion	criterion	NOUN
ajis02-348	234	8	3	3	NUM
ajis02-348	234	9	with	with	ADP
ajis02-348	234	10	respect	respect	NOUN
ajis02-348	234	11	to	to	ADP
ajis02-348	234	12	estimation	estimation	NOUN
ajis02-348	234	13	bias	bias	NOUN
ajis02-348	234	14	,	,	PUNCT
ajis02-348	234	15	because	because	SCONJ
ajis02-348	234	16	the	the	DET
ajis02-348	234	17	actt	actt	NOUN
ajis02-348	234	18	and	and	CCONJ
ajis02-348	234	19	estt	estt	NOUN
ajis02-348	234	20	can	can	AUX
ajis02-348	234	21	interchange	interchange	VERB
ajis02-348	234	22	in	in	ADP
ajis02-348	234	23	the	the	DET
ajis02-348	234	24	expression	expression	NOUN
ajis02-348	234	25	of	of	ADP
ajis02-348	234	26	criterion	criterion	NOUN
ajis02-348	234	27	1	1	NUM
ajis02-348	234	28	.	.	PUNCT
ajis02-348	234	29	)	)	PUNCT
ajis02-348	235	1	this	this	PRON
ajis02-348	235	2	shows	show	VERB
ajis02-348	235	3	that	that	SCONJ
ajis02-348	235	4	the	the	DET
ajis02-348	235	5	regression	regression	NOUN
ajis02-348	235	6	with	with	ADP
ajis02-348	235	7	criterion	criterion	NOUN
ajis02-348	235	8	3	3	NUM
ajis02-348	235	9	overcomes	overcome	VERB
ajis02-348	235	10	the	the	DET
ajis02-348	235	11	shortcoming	shortcoming	NOUN
ajis02-348	235	12	of	of	ADP
ajis02-348	235	13	the	the	DET
ajis02-348	235	14	tendency	tendency	NOUN
ajis02-348	235	15	to	to	PART
ajis02-348	235	16	underestimate	underestimate	VERB
ajis02-348	235	17	which	which	PRON
ajis02-348	235	18	occurred	occur	VERB
ajis02-348	235	19	with	with	ADP
ajis02-348	235	20	criterion	criterion	NOUN
ajis02-348	235	21	2	2	NUM
ajis02-348	235	22	.	.	PUNCT
ajis02-348	235	23	figure	figure	NOUN
ajis02-348	235	24	2	2	NUM
ajis02-348	235	25	relative	relative	ADJ
ajis02-348	235	26	errors	error	NOUN
ajis02-348	235	27	(	(	PUNCT
ajis02-348	235	28	criterion	criterion	NOUN
ajis02-348	235	29	2	2	NUM
ajis02-348	235	30	)	)	PUNCT
ajis02-348	235	31	figure	figure	NOUN
ajis02-348	235	32	3	3	NUM
ajis02-348	235	33	relative	relative	ADJ
ajis02-348	235	34	errors	error	NOUN
ajis02-348	235	35	(	(	PUNCT
ajis02-348	235	36	criterion	criterion	NOUN
ajis02-348	235	37	3	3	NUM
ajis02-348	235	38	)	)	PUNCT
ajis02-348	235	39	assessment	assessment	NOUN
ajis02-348	235	40	on	on	ADP
ajis02-348	235	41	the	the	DET
ajis02-348	235	42	measures	measure	NOUN
ajis02-348	235	43	of	of	ADP
ajis02-348	235	44	accuracy	accuracy	NOUN
ajis02-348	235	45	and	and	CCONJ
ajis02-348	235	46	consistency	consistency	NOUN
ajis02-348	235	47	for	for	ADP
ajis02-348	235	48	comparison	comparison	NOUN
ajis02-348	235	49	,	,	PUNCT
ajis02-348	235	50	q3	q3	PROPN
ajis02-348	235	51	,	,	PUNCT
ajis02-348	235	52	pred(0.25	pred(0.25	NOUN
ajis02-348	235	53	)	)	PUNCT
ajis02-348	235	54	,	,	PUNCT
ajis02-348	235	55	mre	mre	PROPN
ajis02-348	235	56	as	as	ADV
ajis02-348	235	57	well	well	ADV
ajis02-348	235	58	as	as	ADP
ajis02-348	235	59	wmq	wmq	NOUN
ajis02-348	235	60	are	be	AUX
ajis02-348	235	61	computed	compute	VERB
ajis02-348	235	62	as	as	ADP
ajis02-348	235	63	measures	measure	NOUN
ajis02-348	235	64	of	of	ADP
ajis02-348	235	65	prediction	prediction	NOUN
ajis02-348	235	66	accuracy	accuracy	NOUN
ajis02-348	235	67	,	,	PUNCT
ajis02-348	235	68	while	while	SCONJ
ajis02-348	235	69	sdr	sdr	PROPN
ajis02-348	235	70	and	and	CCONJ
ajis02-348	235	71	sdr	sdr	PROPN
ajis02-348	235	72	are	be	AUX
ajis02-348	235	73	computed	compute	VERB
ajis02-348	235	74	as	as	ADP
ajis02-348	235	75	measures	measure	NOUN
ajis02-348	235	76	of	of	ADP
ajis02-348	235	77	estimation	estimation	NOUN
ajis02-348	235	78	consistency	consistency	NOUN
ajis02-348	235	79	.	.	PUNCT
ajis02-348	236	1	we	we	PRON
ajis02-348	236	2	also	also	ADV
ajis02-348	236	3	use	use	VERB
ajis02-348	236	4	ajis	ajis	ADV
ajis02-348	236	5	vol	vol	NOUN
ajis02-348	236	6	5	5	NUM
ajis02-348	236	7	no	no	DET
ajis02-348	236	8	1	1	NUM
ajis02-348	236	9	september	september	NOUN
ajis02-348	236	10	1997	1997	NUM
ajis02-348	236	11	40	40	NUM
ajis02-348	236	12	effort	effort	NOUN
ajis02-348	236	13	=	=	SYM
ajis02-348	236	14	a(fp)b	a(fp)b	NOUN
ajis02-348	236	15	for	for	ADP
ajis02-348	236	16	regression	regression	NOUN
ajis02-348	236	17	in	in	ADP
ajis02-348	236	18	order	order	NOUN
ajis02-348	236	19	to	to	PART
ajis02-348	236	20	fully	fully	ADV
ajis02-348	236	21	assess	assess	VERB
ajis02-348	236	22	the	the	DET
ajis02-348	236	23	measures	measure	NOUN
ajis02-348	236	24	of	of	ADP
ajis02-348	236	25	accuracy	accuracy	NOUN
ajis02-348	236	26	.	.	PUNCT
ajis02-348	237	1	accuracy	accuracy	NOUN
ajis02-348	237	2	assessment	assessment	NOUN
ajis02-348	237	3	table	table	NOUN
ajis02-348	237	4	1	1	NUM
ajis02-348	237	5	uses	use	VERB
ajis02-348	237	6	fp	fp	NOUN
ajis02-348	237	7	as	as	ADP
ajis02-348	237	8	size	size	NOUN
ajis02-348	237	9	measure	measure	NOUN
ajis02-348	237	10	while	while	SCONJ
ajis02-348	237	11	table	table	NOUN
ajis02-348	237	12	2	2	NUM
ajis02-348	237	13	uses	use	VERB
ajis02-348	237	14	wfp	wfp	PROPN
ajis02-348	237	15	.	.	PUNCT
ajis02-348	237	16	table	table	NOUN
ajis02-348	237	17	2	2	NUM
ajis02-348	237	18	also	also	ADV
ajis02-348	237	19	includes	include	VERB
ajis02-348	237	20	consistency	consistency	NOUN
ajis02-348	237	21	results	result	NOUN
ajis02-348	237	22	,	,	PUNCT
ajis02-348	237	23	which	which	PRON
ajis02-348	237	24	are	be	AUX
ajis02-348	237	25	to	to	PART
ajis02-348	237	26	be	be	AUX
ajis02-348	237	27	compared	compare	VERB
ajis02-348	237	28	with	with	ADP
ajis02-348	237	29	table	table	NOUN
ajis02-348	237	30	3	3	NUM
ajis02-348	237	31	in	in	ADP
ajis02-348	237	32	the	the	DET
ajis02-348	237	33	next	next	ADJ
ajis02-348	237	34	section	section	NOUN
ajis02-348	237	35	.	.	PUNCT
ajis02-348	238	1	table	table	NOUN
ajis02-348	238	2	1	1	NUM
ajis02-348	238	3	accuracy	accuracy	NOUN
ajis02-348	238	4	of	of	ADP
ajis02-348	238	5	fp	fp	NOUN
ajis02-348	238	6	as	as	ADP
ajis02-348	238	7	size	size	NOUN
ajis02-348	238	8	measure	measure	NOUN
ajis02-348	238	9	criterion	criterion	NOUN
ajis02-348	238	10	1	1	NUM
ajis02-348	238	11	2	2	NUM
ajis02-348	238	12	3	3	NUM
ajis02-348	238	13	ft	ft	ADP
ajis02-348	238	14	0.66	0.66	NUM
ajis02-348	238	15	0.70	0.70	NUM
ajis02-348	238	16	0.56	0.56	NUM
ajis02-348	238	17	pred(0.25	pred(0.25	NOUN
ajis02-348	238	18	)	)	PUNCT
ajis02-348	238	19	0.42	0.42	NUM
ajis02-348	238	20	0.11	0.11	NUM
ajis02-348	238	21	0.37	0.37	NUM
ajis02-348	238	22	wmq	wmq	NOUN
ajis02-348	238	23	0.45	0.45	NUM
ajis02-348	238	24	0.62	0.62	NUM
ajis02-348	238	25	0.42	0.42	NUM
ajis02-348	238	26	mre	mre	NOUN
ajis02-348	238	27	0.67	0.67	NUM
ajis02-348	238	28	0.58	0.58	NUM
ajis02-348	238	29	0.58	0.58	NUM
ajis02-348	238	30	table	table	NOUN
ajis02-348	238	31	2	2	NUM
ajis02-348	238	32	accuracy	accuracy	NOUN
ajis02-348	238	33	and	and	CCONJ
ajis02-348	238	34	consistency	consistency	NOUN
ajis02-348	238	35	of	of	ADP
ajis02-348	238	36	wfp	wfp	PROPN
ajis02-348	238	37	as	as	ADP
ajis02-348	238	38	size	size	NOUN
ajis02-348	238	39	measure	measure	NOUN
ajis02-348	238	40	criterion	criterion	NOUN
ajis02-348	238	41	1	1	NUM
ajis02-348	238	42	2	2	NUM
ajis02-348	238	43	3	3	NUM
ajis02-348	238	44	ft	ft	NOUN
ajis02-348	238	45	0.43	0.43	NUM
ajis02-348	238	46	0.45	0.45	NUM
ajis02-348	238	47	0.44	0.44	NUM
ajis02-348	238	48	pred(0.25	pred(0.25	NOUN
ajis02-348	238	49	)	)	PUNCT
ajis02-348	238	50	0.49	0.49	NUM
ajis02-348	238	51	0.48	0.48	NUM
ajis02-348	238	52	0.56	0.56	NUM
ajis02-348	238	53	wmq	wmq	NOUN
ajis02-348	238	54	0.32	0.32	NUM
ajis02-348	238	55	0.34	0.34	NUM
ajis02-348	238	56	0.31	0.31	NUM
ajis02-348	238	57	mre	mre	NOUN
ajis02-348	238	58	0.33	0.33	NUM
ajis02-348	238	59	0.30	0.30	NUM
ajis02-348	238	60	0.32	0.32	NUM
ajis02-348	238	61	r	r	NOUN
ajis02-348	238	62	0.8428	0.8428	NUM
ajis02-348	238	63	0.8389	0.8389	NUM
ajis02-348	238	64	0	0	NUM
ajis02-348	238	65	8406	8406	NUM
ajis02-348	238	66	sdr	sdr	PROPN
ajis02-348	238	67	0.439	0.439	NUM
ajis02-348	238	68	0.344	0.344	NUM
ajis02-348	238	69	0.423	0.423	NUM
ajis02-348	238	70	the	the	DET
ajis02-348	238	71	first	first	ADJ
ajis02-348	238	72	observation	observation	NOUN
ajis02-348	238	73	is	be	AUX
ajis02-348	238	74	that	that	SCONJ
ajis02-348	238	75	the	the	DET
ajis02-348	238	76	compared	compare	VERB
ajis02-348	238	77	measures	measure	NOUN
ajis02-348	238	78	of	of	ADP
ajis02-348	238	79	accuracy	accuracy	NOUN
ajis02-348	238	80	,	,	PUNCT
ajis02-348	238	81	q3	q3	PROPN
ajis02-348	238	82	,	,	PUNCT
ajis02-348	238	83	pred(0.25	pred(0.25	NOUN
ajis02-348	238	84	)	)	PUNCT
ajis02-348	238	85	,	,	PUNCT
ajis02-348	238	86	wmq	wmq	NOUN
ajis02-348	238	87	and	and	CCONJ
ajis02-348	238	88	mre	mre	PROPN
ajis02-348	238	89	,	,	PUNCT
ajis02-348	238	90	are	be	AUX
ajis02-348	238	91	not	not	PART
ajis02-348	238	92	consistent	consistent	ADJ
ajis02-348	238	93	in	in	ADP
ajis02-348	238	94	both	both	DET
ajis02-348	238	95	tables	table	NOUN
ajis02-348	238	96	.	.	PUNCT
ajis02-348	239	1	this	this	DET
ajis02-348	239	2	observation	observation	NOUN
ajis02-348	239	3	is	be	AUX
ajis02-348	239	4	not	not	PART
ajis02-348	239	5	unexpected	unexpected	ADJ
ajis02-348	239	6	because	because	SCONJ
ajis02-348	239	7	the	the	DET
ajis02-348	239	8	q^	q^	NOUN
ajis02-348	239	9	and	and	CCONJ
ajis02-348	239	10	pred(0.25	pred(0.25	NOUN
ajis02-348	239	11	)	)	PUNCT
ajis02-348	239	12	are	be	AUX
ajis02-348	239	13	stochastic	stochastic	ADJ
ajis02-348	239	14	.	.	PUNCT
ajis02-348	240	1	the	the	DET
ajis02-348	240	2	difference	difference	NOUN
ajis02-348	240	3	between	between	ADP
ajis02-348	240	4	mre	mre	PROPN
ajis02-348	240	5	and	and	CCONJ
ajis02-348	240	6	wmq	wmq	NOUN
ajis02-348	240	7	will	will	AUX
ajis02-348	240	8	be	be	AUX
ajis02-348	240	9	discussed	discuss	VERB
ajis02-348	240	10	later	later	ADV
ajis02-348	240	11	.	.	PUNCT
ajis02-348	241	1	the	the	DET
ajis02-348	241	2	second	second	ADJ
ajis02-348	241	3	observation	observation	NOUN
ajis02-348	241	4	is	be	AUX
ajis02-348	241	5	that	that	SCONJ
ajis02-348	241	6	criterion	criterion	NOUN
ajis02-348	241	7	2	2	NUM
ajis02-348	241	8	leads	lead	VERB
ajis02-348	241	9	to	to	ADP
ajis02-348	241	10	the	the	DET
ajis02-348	241	11	best	good	ADJ
ajis02-348	241	12	mre	mre	NOUN
ajis02-348	241	13	and	and	CCONJ
ajis02-348	241	14	criterion	criterion	NOUN
ajis02-348	241	15	1	1	NUM
ajis02-348	241	16	leads	lead	VERB
ajis02-348	241	17	to	to	ADP
ajis02-348	241	18	the	the	DET
ajis02-348	241	19	worst	bad	ADJ
ajis02-348	241	20	mre	mre	NOUN
ajis02-348	241	21	.	.	PROPN
ajis02-348	241	22	as	as	SCONJ
ajis02-348	241	23	discussed	discuss	VERB
ajis02-348	241	24	in	in	ADP
ajis02-348	241	25	the	the	DET
ajis02-348	241	26	previous	previous	ADJ
ajis02-348	241	27	section	section	NOUN
ajis02-348	241	28	,	,	PUNCT
ajis02-348	241	29	criterion	criterion	NOUN
ajis02-348	241	30	1	1	NUM
ajis02-348	241	31	is	be	AUX
ajis02-348	241	32	based	base	VERB
ajis02-348	241	33	on	on	ADP
ajis02-348	241	34	difference	difference	NOUN
ajis02-348	241	35	measure	measure	NOUN
ajis02-348	241	36	while	while	SCONJ
ajis02-348	241	37	mre	mre	PROPN
ajis02-348	241	38	is	be	AUX
ajis02-348	241	39	a	a	DET
ajis02-348	241	40	ratio	ratio	NOUN
ajis02-348	241	41	measure	measure	NOUN
ajis02-348	241	42	.	.	PUNCT
ajis02-348	242	1	therefore	therefore	ADV
ajis02-348	242	2	the	the	DET
ajis02-348	242	3	mre	mre	NOUN
ajis02-348	242	4	of	of	ADP
ajis02-348	242	5	criterion	criterion	NOUN
ajis02-348	242	6	1	1	NUM
ajis02-348	242	7	is	be	AUX
ajis02-348	242	8	larger	large	ADJ
ajis02-348	242	9	than	than	ADP
ajis02-348	242	10	that	that	PRON
ajis02-348	242	11	of	of	ADP
ajis02-348	242	12	criterion	criterion	NOUN
ajis02-348	242	13	2	2	NUM
ajis02-348	242	14	and	and	CCONJ
ajis02-348	242	15	criterion	criterion	NOUN
ajis02-348	242	16	3	3	NUM
ajis02-348	242	17	,	,	PUNCT
ajis02-348	242	18	which	which	PRON
ajis02-348	242	19	are	be	AUX
ajis02-348	242	20	based	base	VERB
ajis02-348	242	21	on	on	ADP
ajis02-348	242	22	relative	relative	ADJ
ajis02-348	242	23	errors	error	NOUN
ajis02-348	242	24	.	.	PUNCT
ajis02-348	243	1	criterion	criterion	NOUN
ajis02-348	243	2	2	2	NUM
ajis02-348	243	3	leads	lead	VERB
ajis02-348	243	4	to	to	ADP
ajis02-348	243	5	underestimation	underestimation	NOUN
ajis02-348	243	6	tendency	tendency	NOUN
ajis02-348	243	7	and	and	CCONJ
ajis02-348	243	8	mre	mre	PROPN
ajis02-348	243	9	favours	favour	VERB
ajis02-348	243	10	underestimation	underestimation	NOUN
ajis02-348	243	11	.	.	PUNCT
ajis02-348	244	1	therefore	therefore	ADV
ajis02-348	244	2	criterion	criterion	NOUN
ajis02-348	244	3	2	2	NUM
ajis02-348	244	4	has	have	VERB
ajis02-348	244	5	the	the	DET
ajis02-348	244	6	best	good	ADJ
ajis02-348	244	7	mre	mre	PROPN
ajis02-348	244	8	.	.	PROPN
ajis02-348	244	9	criterion	criterion	NOUN
ajis02-348	244	10	3	3	NUM
ajis02-348	244	11	does	do	AUX
ajis02-348	244	12	not	not	PART
ajis02-348	244	13	have	have	VERB
ajis02-348	244	14	overor	overor	NOUN
ajis02-348	244	15	underestimation	underestimation	NOUN
ajis02-348	244	16	tendency	tendency	NOUN
ajis02-348	244	17	.	.	PUNCT
ajis02-348	245	1	therefore	therefore	ADV
ajis02-348	245	2	the	the	DET
ajis02-348	245	3	mre	mre	NOUN
ajis02-348	245	4	of	of	ADP
ajis02-348	245	5	criterion	criterion	NOUN
ajis02-348	245	6	3	3	NUM
ajis02-348	245	7	is	be	AUX
ajis02-348	245	8	located	locate	VERB
ajis02-348	245	9	between	between	ADP
ajis02-348	245	10	those	those	PRON
ajis02-348	245	11	of	of	ADP
ajis02-348	245	12	criterion	criterion	NOUN
ajis02-348	245	13	2	2	NUM
ajis02-348	245	14	and	and	CCONJ
ajis02-348	245	15	criterion	criterion	NOUN
ajis02-348	245	16	1	1	NUM
ajis02-348	245	17	.	.	PUNCT
ajis02-348	246	1	the	the	DET
ajis02-348	246	2	third	third	ADJ
ajis02-348	246	3	observation	observation	NOUN
ajis02-348	246	4	is	be	AUX
ajis02-348	246	5	that	that	SCONJ
ajis02-348	246	6	if	if	SCONJ
ajis02-348	246	7	there	there	PRON
ajis02-348	246	8	are	be	VERB
ajis02-348	246	9	no	no	DET
ajis02-348	246	10	outliers	outlier	NOUN
ajis02-348	246	11	and	and	CCONJ
ajis02-348	246	12	no	no	DET
ajis02-348	246	13	underestimation	underestimation	NOUN
ajis02-348	246	14	tendency	tendency	NOUN
ajis02-348	246	15	(	(	PUNCT
ajis02-348	246	16	criterions	criterion	NOUN
ajis02-348	246	17	1	1	NUM
ajis02-348	246	18	and	and	CCONJ
ajis02-348	246	19	3	3	NUM
ajis02-348	246	20	in	in	ADP
ajis02-348	246	21	table	table	NOUN
ajis02-348	246	22	2	2	NUM
ajis02-348	246	23	)	)	PUNCT
ajis02-348	246	24	,	,	PUNCT
ajis02-348	246	25	wmq	wmq	NOUN
ajis02-348	246	26	=	=	SYM
ajis02-348	246	27	mre	mre	PROPN
ajis02-348	246	28	.	.	PUNCT
ajis02-348	247	1	if	if	SCONJ
ajis02-348	247	2	there	there	PRON
ajis02-348	247	3	are	be	VERB
ajis02-348	247	4	outliers	outlier	NOUN
ajis02-348	247	5	(	(	PUNCT
ajis02-348	247	6	criterion	criterion	NOUN
ajis02-348	247	7	1	1	NUM
ajis02-348	247	8	and	and	CCONJ
ajis02-348	247	9	criterion	criterion	NOUN
ajis02-348	247	10	3	3	NUM
ajis02-348	247	11	in	in	ADP
ajis02-348	247	12	table	table	NOUN
ajis02-348	247	13	1	1	NUM
ajis02-348	247	14	)	)	PUNCT
ajis02-348	247	15	,	,	PUNCT
ajis02-348	247	16	wmq	wmq	ADP
ajis02-348	247	17	<	<	X
ajis02-348	247	18	mre	mre	PROPN
ajis02-348	247	19	,	,	PUNCT
ajis02-348	247	20	and	and	CCONJ
ajis02-348	247	21	if	if	SCONJ
ajis02-348	247	22	underestimation	underestimation	ADJ
ajis02-348	247	23	tendency	tendency	NOUN
ajis02-348	247	24	exists	exist	VERB
ajis02-348	247	25	(	(	PUNCT
ajis02-348	247	26	criterion	criterion	NOUN
ajis02-348	247	27	2	2	NUM
ajis02-348	247	28	in	in	ADP
ajis02-348	247	29	table	table	NOUN
ajis02-348	247	30	1	1	NUM
ajis02-348	247	31	)	)	PUNCT
ajis02-348	247	32	,	,	PUNCT
ajis02-348	247	33	wmq	wmq	PROPN
ajis02-348	247	34	>	>	X
ajis02-348	247	35	mre	mre	PROPN
ajis02-348	247	36	.	.	PROPN
ajis02-348	248	1	as	as	SCONJ
ajis02-348	248	2	mentioned	mention	VERB
ajis02-348	248	3	before	before	ADV
ajis02-348	248	4	,	,	PUNCT
ajis02-348	248	5	mre	mre	PROPN
ajis02-348	248	6	as	as	ADP
ajis02-348	248	7	a	a	DET
ajis02-348	248	8	measure	measure	NOUN
ajis02-348	248	9	of	of	ADP
ajis02-348	248	10	accuracy	accuracy	NOUN
ajis02-348	248	11	has	have	VERB
ajis02-348	248	12	two	two	NUM
ajis02-348	248	13	drawbacks	drawback	NOUN
ajis02-348	248	14	.	.	PUNCT
ajis02-348	249	1	it	it	PRON
ajis02-348	249	2	is	be	AUX
ajis02-348	249	3	influenced	influence	VERB
ajis02-348	249	4	by	by	ADP
ajis02-348	249	5	outliers	outlier	NOUN
ajis02-348	249	6	and	and	CCONJ
ajis02-348	249	7	it	it	PRON
ajis02-348	249	8	favours	favour	VERB
ajis02-348	249	9	underestimation	underestimation	NOUN
ajis02-348	249	10	.	.	PUNCT
ajis02-348	250	1	wmq	wmq	NOUN
ajis02-348	250	2	as	as	ADP
ajis02-348	250	3	proposed	propose	VERB
ajis02-348	250	4	targets	target	NOUN
ajis02-348	250	5	the	the	DET
ajis02-348	250	6	solution	solution	NOUN
ajis02-348	250	7	of	of	ADP
ajis02-348	250	8	these	these	DET
ajis02-348	250	9	two	two	NUM
ajis02-348	250	10	difficulties	difficulty	NOUN
ajis02-348	250	11	,	,	PUNCT
ajis02-348	250	12	and	and	CCONJ
ajis02-348	250	13	this	this	DET
ajis02-348	250	14	third	third	ADJ
ajis02-348	250	15	observation	observation	NOUN
ajis02-348	250	16	indicates	indicate	VERB
ajis02-348	250	17	that	that	SCONJ
ajis02-348	250	18	wmq	wmq	NOUN
ajis02-348	250	19	has	have	AUX
ajis02-348	250	20	succeeded	succeed	VERB
ajis02-348	250	21	in	in	ADP
ajis02-348	250	22	its	its	PRON
ajis02-348	250	23	objective	objective	NOUN
ajis02-348	250	24	.	.	PUNCT
ajis02-348	251	1	the	the	DET
ajis02-348	251	2	fourth	fourth	ADJ
ajis02-348	251	3	observation	observation	NOUN
ajis02-348	251	4	is	be	AUX
ajis02-348	251	5	that	that	SCONJ
ajis02-348	251	6	the	the	DET
ajis02-348	251	7	criterion	criterion	NOUN
ajis02-348	251	8	3	3	NUM
ajis02-348	251	9	leads	lead	VERB
ajis02-348	251	10	to	to	ADP
ajis02-348	251	11	the	the	DET
ajis02-348	251	12	best	good	ADJ
ajis02-348	251	13	wmq	wmq	NOUN
ajis02-348	251	14	for	for	ADP
ajis02-348	251	15	all	all	DET
ajis02-348	251	16	data	datum	NOUN
ajis02-348	251	17	conditions	condition	NOUN
ajis02-348	251	18	.	.	PUNCT
ajis02-348	252	1	as	as	SCONJ
ajis02-348	252	2	assessed	assess	VERB
ajis02-348	252	3	above	above	ADV
ajis02-348	252	4	,	,	PUNCT
ajis02-348	252	5	wmq	wmq	PROPN
ajis02-348	252	6	is	be	AUX
ajis02-348	252	7	a	a	DET
ajis02-348	252	8	better	well	ADJ
ajis02-348	252	9	measure	measure	NOUN
ajis02-348	252	10	of	of	ADP
ajis02-348	252	11	accuracy	accuracy	NOUN
ajis02-348	252	12	,	,	PUNCT
ajis02-348	252	13	so	so	CCONJ
ajis02-348	252	14	criterion	criterion	NOUN
ajis02-348	252	15	3	3	NUM
ajis02-348	252	16	is	be	AUX
ajis02-348	252	17	superior	superior	ADJ
ajis02-348	252	18	to	to	ADP
ajis02-348	252	19	the	the	DET
ajis02-348	252	20	other	other	ADJ
ajis02-348	252	21	two	two	NUM
ajis02-348	252	22	criteria	criterion	NOUN
ajis02-348	252	23	.	.	PUNCT
ajis02-348	253	1	consistency	consistency	NOUN
ajis02-348	253	2	assessment	assessment	NOUN
ajis02-348	253	3	in	in	ADP
ajis02-348	253	4	this	this	DET
ajis02-348	253	5	section	section	NOUN
ajis02-348	253	6	,	,	PUNCT
ajis02-348	253	7	it	it	PRON
ajis02-348	253	8	will	will	AUX
ajis02-348	253	9	be	be	AUX
ajis02-348	253	10	first	first	ADV
ajis02-348	253	11	shown	show	VERB
ajis02-348	253	12	that	that	SCONJ
ajis02-348	253	13	sdr	sdr	PROPN
ajis02-348	253	14	is	be	AUX
ajis02-348	253	15	better	well	ADJ
ajis02-348	253	16	than	than	ADP
ajis02-348	253	17	sdr	sdr	PROPN
ajis02-348	253	18	,	,	PUNCT
ajis02-348	253	19	because	because	SCONJ
ajis02-348	253	20	r	r	NOUN
ajis02-348	253	21	is	be	AUX
ajis02-348	253	22	influenced	influence	VERB
ajis02-348	253	23	by	by	ADP
ajis02-348	253	24	not	not	PART
ajis02-348	253	25	only	only	ADV
ajis02-348	253	26	the	the	DET
ajis02-348	253	27	estimation	estimation	NOUN
ajis02-348	253	28	result	result	NOUN
ajis02-348	253	29	but	but	CCONJ
ajis02-348	253	30	also	also	ADV
ajis02-348	253	31	the	the	DET
ajis02-348	253	32	variance	variance	NOUN
ajis02-348	253	33	of	of	ADP
ajis02-348	253	34	the	the	DET
ajis02-348	253	35	actual	actual	ADJ
ajis02-348	253	36	effort	effort	NOUN
ajis02-348	253	37	,	,	PUNCT
ajis02-348	253	38	whereas	whereas	SCONJ
ajis02-348	253	39	sdr	sdr	PROPN
ajis02-348	253	40	is	be	AUX
ajis02-348	253	41	not	not	PART
ajis02-348	253	42	influenced	influence	VERB
ajis02-348	253	43	by	by	ADP
ajis02-348	253	44	the	the	DET
ajis02-348	253	45	later	later	NOUN
ajis02-348	253	46	.	.	PUNCT
ajis02-348	254	1	for	for	ADP
ajis02-348	254	2	this	this	DET
ajis02-348	254	3	purpose	purpose	NOUN
ajis02-348	254	4	,	,	PUNCT
ajis02-348	254	5	the	the	DET
ajis02-348	254	6	data	datum	NOUN
ajis02-348	254	7	of	of	ADP
ajis02-348	254	8	one	one	NUM
ajis02-348	254	9	project	project	NOUN
ajis02-348	254	10	,	,	PUNCT
ajis02-348	254	11	which	which	PRON
ajis02-348	254	12	involves	involve	VERB
ajis02-348	254	13	the	the	DET
ajis02-348	254	14	largest	large	ADJ
ajis02-348	254	15	development	development	NOUN
ajis02-348	254	16	effort	effort	NOUN
ajis02-348	254	17	,	,	PUNCT
ajis02-348	254	18	is	be	AUX
ajis02-348	254	19	deleted	delete	VERB
ajis02-348	254	20	from	from	ADP
ajis02-348	254	21	the	the	DET
ajis02-348	254	22	data	datum	NOUN
ajis02-348	254	23	set	set	VERB
ajis02-348	254	24	.	.	PUNCT
ajis02-348	255	1	with	with	ADP
ajis02-348	255	2	aj1s	aj1s	PROPN
ajis02-348	255	3	vol5	vol5	NOUN
ajis02-348	255	4	no	no	DET
ajis02-348	255	5	1	1	NUM
ajis02-348	255	6	september	september	NOUN
ajis02-348	255	7	1997	1997	NUM
ajis02-348	255	8	41	41	NUM
ajis02-348	255	9	this	this	DET
ajis02-348	255	10	change	change	NOUN
ajis02-348	255	11	the	the	DET
ajis02-348	255	12	variance	variance	NOUN
ajis02-348	255	13	of	of	ADP
ajis02-348	255	14	the	the	DET
ajis02-348	255	15	actual	actual	ADJ
ajis02-348	255	16	effort	effort	NOUN
ajis02-348	255	17	changes	change	NOUN
ajis02-348	255	18	.	.	PUNCT
ajis02-348	256	1	table	table	NOUN
ajis02-348	256	2	3	3	NUM
ajis02-348	256	3	shows	show	VERB
ajis02-348	256	4	the	the	DET
ajis02-348	256	5	accuracy	accuracy	NOUN
ajis02-348	256	6	and	and	CCONJ
ajis02-348	256	7	consistency	consistency	NOUN
ajis02-348	256	8	results	result	NOUN
ajis02-348	256	9	from	from	ADP
ajis02-348	256	10	the	the	DET
ajis02-348	256	11	reduced	reduce	VERB
ajis02-348	256	12	data	datum	NOUN
ajis02-348	256	13	set	set	VERB
ajis02-348	256	14	.	.	PUNCT
ajis02-348	257	1	table	table	NOUN
ajis02-348	257	2	3	3	NUM
ajis02-348	257	3	accuracy	accuracy	NOUN
ajis02-348	257	4	and	and	CCONJ
ajis02-348	257	5	consistency	consistency	NOUN
ajis02-348	257	6	results	result	NOUN
ajis02-348	257	7	from	from	ADP
ajis02-348	257	8	reduced	reduce	VERB
ajis02-348	257	9	data	datum	NOUN
ajis02-348	257	10	set	set	VERB
ajis02-348	257	11	criterion	criterion	NOUN
ajis02-348	257	12	q3	q3	PROPN
ajis02-348	257	13	pred(0.25	pred(0.25	NOUN
ajis02-348	257	14	)	)	PUNCT
ajis02-348	257	15	wmq	wmq	NOUN
ajis02-348	257	16	1	1	NUM
ajis02-348	257	17	2	2	NUM
ajis02-348	257	18	3	3	NUM
ajis02-348	257	19	0.44	0.44	NUM
ajis02-348	257	20	0.45	0.45	NUM
ajis02-348	257	21	0.44	0.44	NUM
ajis02-348	257	22	0.49	0.49	NUM
ajis02-348	257	23	0.49	0.49	NUM
ajis02-348	257	24	0.56	0.56	NUM
ajis02-348	257	25	0.33	0.33	NUM
ajis02-348	257	26	0.33	0.33	NUM
ajis02-348	257	27	0.31	0.31	NUM
ajis02-348	257	28	mre	mre	NOUN
ajis02-348	257	29	0.34	0.34	NUM
ajis02-348	258	1	0.30	0.30	NUM
ajis02-348	258	2	0.32	0.32	NUM
ajis02-348	258	3	r	r	NOUN
ajis02-348	258	4	0.7900	0.7900	NUM
ajis02-348	258	5	0.7885	0.7885	NUM
ajis02-348	258	6	0.7891	0.7891	NUM
ajis02-348	258	7	sdr	sdr	NOUN
ajis02-348	258	8	0.441	0.441	NUM
ajis02-348	258	9	0.345	0.345	NUM
ajis02-348	258	10	0.425	0.425	NUM
ajis02-348	258	11	comparing	compare	VERB
ajis02-348	258	12	table	table	NOUN
ajis02-348	258	13	2	2	NUM
ajis02-348	258	14	and	and	CCONJ
ajis02-348	258	15	table	table	NOUN
ajis02-348	258	16	3	3	NUM
ajis02-348	258	17	,	,	PUNCT
ajis02-348	258	18	we	we	PRON
ajis02-348	258	19	find	find	VERB
ajis02-348	258	20	that	that	SCONJ
ajis02-348	258	21	the	the	DET
ajis02-348	258	22	q	q	NOUN
ajis02-348	258	23	values	value	NOUN
ajis02-348	258	24	,	,	PUNCT
ajis02-348	258	25	wmq	wmq	NOUN
ajis02-348	258	26	and	and	CCONJ
ajis02-348	258	27	mre	mre	NOUN
ajis02-348	258	28	of	of	ADP
ajis02-348	258	29	each	each	DET
ajis02-348	258	30	criterion	criterion	NOUN
ajis02-348	258	31	are	be	AUX
ajis02-348	258	32	virtually	virtually	ADV
ajis02-348	258	33	the	the	DET
ajis02-348	258	34	same	same	ADJ
ajis02-348	258	35	in	in	ADP
ajis02-348	258	36	the	the	DET
ajis02-348	258	37	two	two	NUM
ajis02-348	258	38	corresponding	corresponding	ADJ
ajis02-348	258	39	situations	situation	NOUN
ajis02-348	258	40	.	.	PUNCT
ajis02-348	259	1	the	the	DET
ajis02-348	259	2	consistency	consistency	NOUN
ajis02-348	259	3	should	should	AUX
ajis02-348	259	4	not	not	PART
ajis02-348	259	5	be	be	AUX
ajis02-348	259	6	different	different	ADJ
ajis02-348	259	7	because	because	SCONJ
ajis02-348	259	8	it	it	PRON
ajis02-348	259	9	measures	measure	VERB
ajis02-348	259	10	the	the	DET
ajis02-348	259	11	quality	quality	NOUN
ajis02-348	259	12	of	of	ADP
ajis02-348	259	13	a	a	DET
ajis02-348	259	14	model	model	NOUN
ajis02-348	259	15	in	in	ADP
ajis02-348	259	16	a	a	DET
ajis02-348	259	17	specific	specific	ADJ
ajis02-348	259	18	environment	environment	NOUN
ajis02-348	259	19	and	and	CCONJ
ajis02-348	259	20	it	it	PRON
ajis02-348	259	21	should	should	AUX
ajis02-348	259	22	be	be	AUX
ajis02-348	259	23	independent	independent	ADJ
ajis02-348	259	24	of	of	ADP
ajis02-348	259	25	the	the	DET
ajis02-348	259	26	variance	variance	NOUN
ajis02-348	259	27	of	of	ADP
ajis02-348	259	28	the	the	DET
ajis02-348	259	29	actual	actual	ADJ
ajis02-348	259	30	effort	effort	NOUN
ajis02-348	259	31	.	.	PUNCT
ajis02-348	260	1	however	however	ADV
ajis02-348	260	2	,	,	PUNCT
ajis02-348	260	3	the	the	DET
ajis02-348	260	4	change	change	NOUN
ajis02-348	260	5	of	of	ADP
ajis02-348	260	6	variance	variance	NOUN
ajis02-348	260	7	makes	make	VERB
ajis02-348	260	8	the	the	DET
ajis02-348	260	9	r	r	NOUN
ajis02-348	260	10	values	value	NOUN
ajis02-348	260	11	decrease	decrease	NOUN
ajis02-348	260	12	by	by	ADP
ajis02-348	260	13	more	more	ADJ
ajis02-348	260	14	than	than	ADP
ajis02-348	260	15	0.0500	0.0500	NUM
ajis02-348	260	16	.	.	PUNCT
ajis02-348	261	1	on	on	ADP
ajis02-348	261	2	the	the	DET
ajis02-348	261	3	other	other	ADJ
ajis02-348	261	4	hand	hand	NOUN
ajis02-348	261	5	,	,	PUNCT
ajis02-348	261	6	the	the	DET
ajis02-348	261	7	sdr	sdr	PROPN
ajis02-348	261	8	values	value	VERB
ajis02-348	261	9	only	only	ADV
ajis02-348	261	10	change	change	VERB
ajis02-348	261	11	by	by	ADP
ajis02-348	261	12	less	less	ADJ
ajis02-348	261	13	than	than	ADP
ajis02-348	261	14	0.002	0.002	NUM
ajis02-348	261	15	,	,	PUNCT
ajis02-348	261	16	which	which	PRON
ajis02-348	261	17	is	be	AUX
ajis02-348	261	18	comparatively	comparatively	ADV
ajis02-348	261	19	small	small	ADJ
ajis02-348	261	20	.	.	PUNCT
ajis02-348	262	1	therefore	therefore	ADV
ajis02-348	262	2	,	,	PUNCT
ajis02-348	262	3	the	the	DET
ajis02-348	262	4	sdr	sdr	PROPN
ajis02-348	262	5	is	be	AUX
ajis02-348	262	6	superior	superior	ADJ
ajis02-348	262	7	over	over	ADP
ajis02-348	262	8	r	r	NOUN
ajis02-348	262	9	in	in	ADP
ajis02-348	262	10	the	the	DET
ajis02-348	262	11	aspect	aspect	NOUN
ajis02-348	262	12	that	that	SCONJ
ajis02-348	262	13	the	the	DET
ajis02-348	262	14	former	former	ADJ
ajis02-348	262	15	is	be	AUX
ajis02-348	262	16	not	not	PART
ajis02-348	262	17	influenced	influence	VERB
ajis02-348	262	18	by	by	ADP
ajis02-348	262	19	the	the	DET
ajis02-348	262	20	variance	variance	NOUN
ajis02-348	262	21	in	in	ADP
ajis02-348	262	22	effort	effort	NOUN
ajis02-348	262	23	.	.	PUNCT
ajis02-348	263	1	next	next	ADV
ajis02-348	263	2	,	,	PUNCT
ajis02-348	263	3	it	it	PRON
ajis02-348	263	4	will	will	AUX
ajis02-348	263	5	be	be	AUX
ajis02-348	263	6	demonstrated	demonstrate	VERB
ajis02-348	263	7	that	that	SCONJ
ajis02-348	263	8	a	a	DET
ajis02-348	263	9	model	model	NOUN
ajis02-348	263	10	producing	produce	VERB
ajis02-348	263	11	consistent	consistent	ADJ
ajis02-348	263	12	estimation	estimation	NOUN
ajis02-348	263	13	errors	error	NOUN
ajis02-348	263	14	is	be	AUX
ajis02-348	263	15	more	more	ADV
ajis02-348	263	16	easily	easily	ADV
ajis02-348	263	17	calibrated	calibrate	VERB
ajis02-348	263	18	to	to	PART
ajis02-348	263	19	improve	improve	VERB
ajis02-348	263	20	accuracy	accuracy	NOUN
ajis02-348	263	21	than	than	SCONJ
ajis02-348	263	22	is	be	AUX
ajis02-348	263	23	a	a	DET
ajis02-348	263	24	model	model	NOUN
ajis02-348	263	25	producing	produce	VERB
ajis02-348	263	26	less	less	ADV
ajis02-348	263	27	consistent	consistent	ADJ
ajis02-348	263	28	errors	error	NOUN
ajis02-348	263	29	,	,	PUNCT
ajis02-348	263	30	if	if	SCONJ
ajis02-348	263	31	the	the	DET
ajis02-348	263	32	consistency	consistency	NOUN
ajis02-348	263	33	is	be	AUX
ajis02-348	263	34	measured	measure	VERB
ajis02-348	263	35	by	by	ADP
ajis02-348	263	36	sdr	sdr	PROPN
ajis02-348	263	37	.	.	PUNCT
ajis02-348	264	1	the	the	DET
ajis02-348	264	2	calibration	calibration	NOUN
ajis02-348	264	3	method	method	NOUN
ajis02-348	264	4	is	be	AUX
ajis02-348	264	5	described	describe	VERB
ajis02-348	264	6	as	as	ADP
ajis02-348	264	7	follows	follow	VERB
ajis02-348	264	8	.	.	PUNCT
ajis02-348	265	1	with	with	ADP
ajis02-348	265	2	desharnais	desharnais	PROPN
ajis02-348	265	3	'	'	PART
ajis02-348	265	4	data	datum	NOUN
ajis02-348	265	5	set	set	VERB
ajis02-348	265	6	,	,	PUNCT
ajis02-348	265	7	three	three	NUM
ajis02-348	265	8	formulas	formula	NOUN
ajis02-348	265	9	are	be	AUX
ajis02-348	265	10	obtained	obtain	VERB
ajis02-348	265	11	by	by	ADP
ajis02-348	265	12	regression	regression	NOUN
ajis02-348	265	13	with	with	ADP
ajis02-348	265	14	the	the	DET
ajis02-348	265	15	three	three	NUM
ajis02-348	265	16	criteria	criterion	NOUN
ajis02-348	265	17	described	describe	VERB
ajis02-348	265	18	earlier	early	ADV
ajis02-348	265	19	.	.	PUNCT
ajis02-348	266	1	these	these	DET
ajis02-348	266	2	formulas	formula	NOUN
ajis02-348	266	3	are	be	AUX
ajis02-348	266	4	to	to	PART
ajis02-348	266	5	be	be	AUX
ajis02-348	266	6	calibrated	calibrate	VERB
ajis02-348	266	7	to	to	ADP
ajis02-348	266	8	the	the	DET
ajis02-348	266	9	environment	environment	NOUN
ajis02-348	266	10	of	of	ADP
ajis02-348	266	11	albrecht	albrecht	PROPN
ajis02-348	266	12	's	's	PART
ajis02-348	266	13	data	datum	NOUN
ajis02-348	266	14	set	set	VERB
ajis02-348	266	15	(	(	PUNCT
ajis02-348	266	16	1983	1983	NUM
ajis02-348	266	17	)	)	PUNCT
ajis02-348	266	18	.	.	PUNCT
ajis02-348	267	1	in	in	ADP
ajis02-348	267	2	albrecht	albrecht	PROPN
ajis02-348	267	3	's	's	PART
ajis02-348	267	4	data	datum	NOUN
ajis02-348	267	5	set	set	VERB
ajis02-348	267	6	,	,	PUNCT
ajis02-348	267	7	15	15	NUM
ajis02-348	267	8	projects	project	NOUN
ajis02-348	267	9	used	use	VERB
ajis02-348	267	10	level	level	NOUN
ajis02-348	267	11	1	1	NUM
ajis02-348	267	12	language	language	NOUN
ajis02-348	267	13	and	and	CCONJ
ajis02-348	267	14	their	their	PRON
ajis02-348	267	15	fps	fps	NOUN
ajis02-348	267	16	fell	fall	VERB
ajis02-348	267	17	in	in	ADP
ajis02-348	267	18	the	the	DET
ajis02-348	267	19	size	size	NOUN
ajis02-348	267	20	range	range	NOUN
ajis02-348	267	21	of	of	ADP
ajis02-348	267	22	desharnais	desharnais	PROPN
ajis02-348	267	23	'	'	PART
ajis02-348	267	24	data	datum	NOUN
ajis02-348	267	25	set	set	VERB
ajis02-348	267	26	.	.	PUNCT
ajis02-348	268	1	for	for	ADP
ajis02-348	268	2	the	the	DET
ajis02-348	268	3	other	other	ADJ
ajis02-348	268	4	projects	project	NOUN
ajis02-348	268	5	,	,	PUNCT
ajis02-348	268	6	either	either	CCONJ
ajis02-348	268	7	the	the	DET
ajis02-348	268	8	languages	language	NOUN
ajis02-348	268	9	can	can	AUX
ajis02-348	268	10	not	not	PART
ajis02-348	268	11	be	be	AUX
ajis02-348	268	12	categorised	categorise	VERB
ajis02-348	268	13	into	into	ADP
ajis02-348	268	14	the	the	DET
ajis02-348	268	15	3	3	NUM
ajis02-348	268	16	levels	level	NOUN
ajis02-348	268	17	or	or	CCONJ
ajis02-348	268	18	the	the	DET
ajis02-348	268	19	fps	fps	NOUN
ajis02-348	268	20	are	be	AUX
ajis02-348	268	21	beyond	beyond	ADP
ajis02-348	268	22	the	the	DET
ajis02-348	268	23	size	size	NOUN
ajis02-348	268	24	range	range	NOUN
ajis02-348	268	25	of	of	ADP
ajis02-348	268	26	desharnais	desharnais	PROPN
ajis02-348	268	27	'	'	PART
ajis02-348	268	28	data	datum	NOUN
ajis02-348	268	29	set	set	NOUN
ajis02-348	268	30	.	.	PUNCT
ajis02-348	269	1	therefore	therefore	ADV
ajis02-348	269	2	only	only	ADV
ajis02-348	269	3	these	these	DET
ajis02-348	269	4	15	15	NUM
ajis02-348	269	5	projects	project	NOUN
ajis02-348	269	6	were	be	AUX
ajis02-348	269	7	used	use	VERB
ajis02-348	269	8	for	for	ADP
ajis02-348	269	9	the	the	DET
ajis02-348	269	10	calibration	calibration	NOUN
ajis02-348	269	11	.	.	PUNCT
ajis02-348	270	1	it	it	PRON
ajis02-348	270	2	is	be	AUX
ajis02-348	270	3	assumed	assume	VERB
ajis02-348	270	4	that	that	SCONJ
ajis02-348	270	5	the	the	DET
ajis02-348	270	6	responsiveness	responsiveness	NOUN
ajis02-348	270	7	of	of	ADP
ajis02-348	270	8	effort	effort	NOUN
ajis02-348	270	9	to	to	PART
ajis02-348	270	10	size	size	VERB
ajis02-348	270	11	(	(	PUNCT
ajis02-348	270	12	exponents	exponent	NOUN
ajis02-348	270	13	in	in	ADP
ajis02-348	270	14	the	the	DET
ajis02-348	270	15	formulas	formula	NOUN
ajis02-348	270	16	,	,	PUNCT
ajis02-348	270	17	refer	refer	VERB
ajis02-348	270	18	to	to	ADP
ajis02-348	270	19	section	section	NOUN
ajis02-348	270	20	of	of	ADP
ajis02-348	270	21	"	"	PUNCT
ajis02-348	270	22	integrated	integrated	ADJ
ajis02-348	270	23	software	software	NOUN
ajis02-348	270	24	cost	cost	NOUN
ajis02-348	270	25	model	model	NOUN
ajis02-348	270	26	"	"	PUNCT
ajis02-348	270	27	)	)	PUNCT
ajis02-348	270	28	does	do	AUX
ajis02-348	270	29	not	not	PART
ajis02-348	270	30	change	change	VERB
ajis02-348	270	31	,	,	PUNCT
ajis02-348	270	32	but	but	CCONJ
ajis02-348	270	33	the	the	DET
ajis02-348	270	34	environments	environment	NOUN
ajis02-348	270	35	of	of	ADP
ajis02-348	270	36	the	the	DET
ajis02-348	270	37	two	two	NUM
ajis02-348	270	38	sets	set	NOUN
ajis02-348	270	39	of	of	ADP
ajis02-348	270	40	data	datum	NOUN
ajis02-348	270	41	are	be	AUX
ajis02-348	270	42	different	different	ADJ
ajis02-348	270	43	.	.	PUNCT
ajis02-348	271	1	in	in	ADP
ajis02-348	271	2	desharnais	desharnais	PROPN
ajis02-348	271	3	'	'	PART
ajis02-348	271	4	environment	environment	NOUN
ajis02-348	271	5	,	,	PUNCT
ajis02-348	271	6	the	the	DET
ajis02-348	271	7	formulas	formula	NOUN
ajis02-348	271	8	obtained	obtain	VERB
ajis02-348	271	9	with	with	ADP
ajis02-348	271	10	the	the	DET
ajis02-348	271	11	three	three	NUM
ajis02-348	271	12	regression	regression	NOUN
ajis02-348	271	13	criteria	criterion	NOUN
ajis02-348	271	14	are	be	AUX
ajis02-348	271	15	effort	effort	NOUN
ajis02-348	271	16	=	=	PUNCT
ajis02-348	271	17	a	a	DET
ajis02-348	271	18	j(wfp)bi	j(wfp)bi	NOUN
ajis02-348	271	19	a	a	DET
ajis02-348	271	20	=	=	NOUN
ajis02-348	271	21	1,2,3	1,2,3	NUM
ajis02-348	271	22	)	)	PUNCT
ajis02-348	271	23	the	the	DET
ajis02-348	271	24	environment	environment	NOUN
ajis02-348	271	25	is	be	AUX
ajis02-348	271	26	the	the	DET
ajis02-348	271	27	overall	overall	ADJ
ajis02-348	271	28	level	level	NOUN
ajis02-348	271	29	of	of	ADP
ajis02-348	271	30	cost	cost	NOUN
ajis02-348	271	31	drivers	driver	NOUN
ajis02-348	271	32	(	(	PUNCT
ajis02-348	271	33	gao	gao	PROPN
ajis02-348	271	34	et	et	PROPN
ajis02-348	271	35	al	al	PROPN
ajis02-348	271	36	,	,	PUNCT
ajis02-348	271	37	1995	1995	NUM
ajis02-348	271	38	)	)	PUNCT
ajis02-348	271	39	,	,	PUNCT
ajis02-348	271	40	which	which	PRON
ajis02-348	271	41	is	be	AUX
ajis02-348	271	42	expressed	express	VERB
ajis02-348	271	43	by	by	ADP
ajis02-348	271	44	the	the	DET
ajis02-348	271	45	coefficient	coefficient	NOUN
ajis02-348	271	46	a	a	PRON
ajis02-348	271	47	}	}	PUNCT
ajis02-348	271	48	(	(	PUNCT
ajis02-348	271	49	j	j	NOUN
ajis02-348	271	50	=	=	NOUN
ajis02-348	271	51	1,2,3	1,2,3	NUM
ajis02-348	271	52	)	)	PUNCT
ajis02-348	271	53	.	.	PUNCT
ajis02-348	272	1	for	for	ADP
ajis02-348	272	2	albrecht	albrecht	PROPN
ajis02-348	272	3	's	's	PART
ajis02-348	272	4	data	datum	NOUN
ajis02-348	272	5	set	set	NOUN
ajis02-348	272	6	,	,	PUNCT
ajis02-348	272	7	an	an	DET
ajis02-348	272	8	environment	environment	NOUN
ajis02-348	272	9	adjustment	adjustment	NOUN
ajis02-348	272	10	parameter	parameter	NOUN
ajis02-348	272	11	,	,	PUNCT
ajis02-348	272	12	kj	kj	PROPN
ajis02-348	272	13	(	(	PUNCT
ajis02-348	272	14	j	j	PROPN
ajis02-348	272	15	=	=	PROPN
ajis02-348	272	16	1,2,3	1,2,3	NUM
ajis02-348	272	17	)	)	PUNCT
ajis02-348	272	18	,	,	PUNCT
ajis02-348	272	19	is	be	AUX
ajis02-348	272	20	used	use	VERB
ajis02-348	272	21	in	in	ADP
ajis02-348	272	22	the	the	DET
ajis02-348	272	23	above	above	ADJ
ajis02-348	272	24	formulas	formula	NOUN
ajis02-348	272	25	to	to	PART
ajis02-348	272	26	account	account	VERB
ajis02-348	272	27	for	for	ADP
ajis02-348	272	28	the	the	DET
ajis02-348	272	29	environment	environment	NOUN
ajis02-348	272	30	change	change	NOUN
ajis02-348	272	31	effort	effort	NOUN
ajis02-348	272	32	=	=	SYM
ajis02-348	272	33	kjaj	kjaj	NOUN
ajis02-348	272	34	(	(	PUNCT
ajis02-348	272	35	j=	j=	NOUN
ajis02-348	272	36	1,2,3	1,2,3	NUM
ajis02-348	272	37	)	)	PUNCT
ajis02-348	272	38	where	where	SCONJ
ajis02-348	272	39	the	the	DET
ajis02-348	272	40	k	k	X
ajis02-348	272	41	(	(	PUNCT
ajis02-348	272	42	/	/	SYM
ajis02-348	272	43	=	=	SYM
ajis02-348	272	44	1	1	NUM
ajis02-348	272	45	»	»	NOUN
ajis02-348	272	46	2,3	2,3	NUM
ajis02-348	272	47	)	)	PUNCT
ajis02-348	272	48	are	be	AUX
ajis02-348	272	49	determined	determine	VERB
ajis02-348	272	50	by	by	ADP
ajis02-348	272	51	the	the	DET
ajis02-348	272	52	following	follow	VERB
ajis02-348	272	53	procedure	procedure	NOUN
ajis02-348	272	54	.	.	PUNCT
ajis02-348	273	1	firstly	firstly	ADV
ajis02-348	273	2	,	,	PUNCT
ajis02-348	273	3	the	the	DET
ajis02-348	273	4	estimated	estimate	VERB
ajis02-348	273	5	efforts	effort	NOUN
ajis02-348	273	6	est(i	est(i	PROPN
ajis02-348	273	7	=	=	SYM
ajis02-348	273	8	1,2,3,	1,2,3,	NUM
ajis02-348	273	9	...	...	PUNCT
ajis02-348	273	10	,15	,15	PUNCT
ajis02-348	273	11	,	,	PUNCT
ajis02-348	273	12	j	j	PROPN
ajis02-348	273	13	=	=	SYM
ajis02-348	273	14	1,2,3	1,2,3	NUM
ajis02-348	273	15	)	)	PUNCT
ajis02-348	273	16	are	be	AUX
ajis02-348	273	17	obtained	obtain	VERB
ajis02-348	273	18	with	with	ADP
ajis02-348	273	19	the	the	DET
ajis02-348	273	20	formulas	formula	NOUN
ajis02-348	273	21	developed	develop	VERB
ajis02-348	273	22	in	in	ADP
ajis02-348	273	23	desharnais	desharnais	PROPN
ajis02-348	273	24	'	'	PART
ajis02-348	273	25	environment	environment	NOUN
ajis02-348	273	26	.	.	PUNCT
ajis02-348	274	1	secondly	secondly	ADV
ajis02-348	274	2	,	,	PUNCT
ajis02-348	274	3	it	it	PRON
ajis02-348	274	4	is	be	AUX
ajis02-348	274	5	assumed	assume	VERB
ajis02-348	274	6	that	that	SCONJ
ajis02-348	274	7	the	the	DET
ajis02-348	274	8	k	k	PROPN
ajis02-348	274	9	•	•	NOUN
ajis02-348	274	10	(	(	PUNCT
ajis02-348	274	11	j	j	NOUN
ajis02-348	274	12	=	=	SYM
ajis02-348	274	13	1,2,3	1,2,3	NUM
ajis02-348	274	14	)	)	PUNCT
ajis02-348	274	15	satisfy	satisfy	VERB
ajis02-348	274	16	the	the	DET
ajis02-348	274	17	following	follow	VERB
ajis02-348	274	18	equations	equation	NOUN
ajis02-348	274	19	est	est	X
ajis02-348	274	20	..	..	PUNCT
ajis02-348	274	21	act	act	VERB
ajis02-348	274	22	:	:	PUNCT
ajis02-348	275	1	=	=	SYM
ajis02-348	275	2	k	k	NOUN
ajis02-348	275	3	,	,	PUNCT
ajis02-348	275	4	est;s	est;s	NOUN
ajis02-348	275	5	orl	orl	PROPN
ajis02-348	275	6	=	=	SYM
ajis02-348	275	7	k.1	k.1	PROPN
ajis02-348	275	8	act	act	NOUN
ajis02-348	275	9	,	,	PUNCT
ajis02-348	275	10	=	=	NOUN
ajis02-348	275	11	1,2,3,	1,2,3,	NUM
ajis02-348	275	12	...	...	PUNCT
ajis02-348	275	13	,15,7=	,15,7=	PUNCT
ajis02-348	275	14	1,2,3	1,2,3	NUM
ajis02-348	275	15	)	)	PUNCT
ajis02-348	275	16	ajis	ajis	ADV
ajis02-348	275	17	vol	vol	VERB
ajis02-348	275	18	5	5	NUM
ajis02-348	275	19	no	no	DET
ajis02-348	275	20	1	1	NUM
ajis02-348	275	21	september	september	NOUN
ajis02-348	275	22	1997	1997	NUM
ajis02-348	275	23	42	42	NUM
ajis02-348	275	24	where	where	SCONJ
ajis02-348	275	25	act	act	NOUN
ajis02-348	275	26	(	(	PUNCT
ajis02-348	275	27	(	(	PUNCT
ajis02-348	275	28	i	i	NOUN
ajis02-348	275	29	=	=	PUNCT
ajis02-348	275	30	1,2,3,	1,2,3,	NUM
ajis02-348	275	31	...	...	PUNCT
ajis02-348	275	32	,15	,15	PUNCT
ajis02-348	275	33	)	)	PUNCT
ajis02-348	275	34	are	be	AUX
ajis02-348	275	35	the	the	DET
ajis02-348	275	36	actual	actual	ADJ
ajis02-348	275	37	values	value	NOUN
ajis02-348	275	38	of	of	ADP
ajis02-348	275	39	effort	effort	NOUN
ajis02-348	275	40	in	in	ADP
ajis02-348	275	41	albrecht	albrecht	PROPN
ajis02-348	275	42	's	's	PART
ajis02-348	275	43	data	datum	NOUN
ajis02-348	275	44	set	set	VERB
ajis02-348	275	45	.	.	PUNCT
ajis02-348	276	1	thirdly	thirdly	ADV
ajis02-348	276	2	,	,	PUNCT
ajis02-348	276	3	these	these	DET
ajis02-348	276	4	15	15	NUM
ajis02-348	276	5	equations	equation	NOUN
ajis02-348	276	6	are	be	AUX
ajis02-348	276	7	summed	sum	VERB
ajis02-348	276	8	to	to	PART
ajis02-348	276	9	give	give	VERB
ajis02-348	276	10	is	be	AUX
ajis02-348	276	11	act	act	NOUN
ajis02-348	276	12	(	(	PUNCT
ajis02-348	276	13	7=1,2,3	7=1,2,3	X
ajis02-348	276	14	)	)	PUNCT
ajis02-348	276	15	and	and	CCONJ
ajis02-348	276	16	the	the	DET
ajis02-348	276	17	above	above	ADJ
ajis02-348	276	18	expression	expression	NOUN
ajis02-348	276	19	is	be	AUX
ajis02-348	276	20	rearranged	rearrange	VERB
ajis02-348	276	21	so	so	SCONJ
ajis02-348	276	22	that	that	SCONJ
ajis02-348	276	23	the	the	DET
ajis02-348	276	24	kj	kj	PROPN
ajis02-348	276	25	(	(	PUNCT
ajis02-348	276	26	j	j	PROPN
ajis02-348	276	27	=	=	SYM
ajis02-348	276	28	1,2,3	1,2,3	NUM
ajis02-348	276	29	)	)	PUNCT
ajis02-348	276	30	are	be	AUX
ajis02-348	276	31	obtained	obtain	VERB
ajis02-348	276	32	15	15	NUM
ajis02-348	276	33	est	est	X
ajis02-348	276	34	0'=	0'=	NUM
ajis02-348	276	35	u2,3	u2,3	ADJ
ajis02-348	276	36	)	)	PUNCT
ajis02-348	276	37	table	table	NOUN
ajis02-348	276	38	4	4	NUM
ajis02-348	276	39	and	and	CCONJ
ajis02-348	276	40	table	table	NOUN
ajis02-348	276	41	5	5	NUM
ajis02-348	276	42	show	show	VERB
ajis02-348	276	43	the	the	DET
ajis02-348	276	44	results	result	NOUN
ajis02-348	276	45	of	of	ADP
ajis02-348	276	46	accuracy	accuracy	NOUN
ajis02-348	276	47	and	and	CCONJ
ajis02-348	276	48	consistency	consistency	NOUN
ajis02-348	276	49	before	before	ADP
ajis02-348	276	50	and	and	CCONJ
ajis02-348	276	51	after	after	ADP
ajis02-348	276	52	the	the	DET
ajis02-348	276	53	calibration	calibration	NOUN
ajis02-348	276	54	.	.	PUNCT
ajis02-348	277	1	table	table	NOUN
ajis02-348	277	2	4	4	NUM
ajis02-348	277	3	results	result	NOUN
ajis02-348	277	4	before	before	ADP
ajis02-348	277	5	calibration	calibration	NOUN
ajis02-348	277	6	criterion	criterion	NOUN
ajis02-348	277	7	1	1	NUM
ajis02-348	277	8	2	2	NUM
ajis02-348	277	9	3	3	NUM
ajis02-348	277	10	g	g	NOUN
ajis02-348	277	11	,	,	PUNCT
ajis02-348	277	12	0.10	0.10	NUM
ajis02-348	277	13	0.21	0.21	NUM
ajis02-348	277	14	0.11	0.11	NUM
ajis02-348	277	15	g2	g2	PROPN
ajis02-348	277	16	0.18	0.18	NUM
ajis02-348	277	17	0.35	0.35	NUM
ajis02-348	277	18	0.26	0.26	NUM
ajis02-348	277	19	a	a	DET
ajis02-348	277	20	0.49	0.49	NUM
ajis02-348	277	21	0.57	0.57	NUM
ajis02-348	277	22	0.47	0.47	NUM
ajis02-348	277	23	pred(0.25	pred(0.25	NOUN
ajis02-348	277	24	)	)	PUNCT
ajis02-348	277	25	0.60	0.60	NUM
ajis02-348	277	26	0.27	0.27	NUM
ajis02-348	277	27	0.47	0.47	NUM
ajis02-348	277	28	wmq	wmq	NOUN
ajis02-348	277	29	0.32	0.32	NUM
ajis02-348	277	30	0.44	0.44	NUM
ajis02-348	277	31	0.34	0.34	NUM
ajis02-348	277	32	mre	mre	NOUN
ajis02-348	277	33	0.36	0.36	NUM
ajis02-348	277	34	0.42	0.42	NUM
ajis02-348	277	35	0.36	0.36	NUM
ajis02-348	278	1	r	r	NOUN
ajis02-348	278	2	0.5328	0.5328	NUM
ajis02-348	278	3	0.5361	0.5361	NUM
ajis02-348	278	4	0.5352	0.5352	NUM
ajis02-348	278	5	sdr	sdr	NOUN
ajis02-348	278	6	0	0	NUM
ajis02-348	278	7	0	0	NUM
ajis02-348	278	8	0	0	NUM
ajis02-348	278	9	57	57	NUM
ajis02-348	278	10	42	42	NUM
ajis02-348	278	11	53	53	NUM
ajis02-348	278	12	table	table	NOUN
ajis02-348	278	13	5	5	NUM
ajis02-348	278	14	results	result	NOUN
ajis02-348	278	15	after	after	ADP
ajis02-348	278	16	calibration	calibration	NOUN
ajis02-348	278	17	criterion	criterion	NOUN
ajis02-348	278	18	q	q	NOUN
ajis02-348	278	19	}	}	PUNCT
ajis02-348	278	20	q2	q2	NOUN
ajis02-348	278	21	1	1	NUM
ajis02-348	278	22	2	2	NUM
ajis02-348	278	23	3	3	NUM
ajis02-348	278	24	0.10	0.10	NUM
ajis02-348	278	25	0.09	0.09	NUM
ajis02-348	278	26	0.10	0.10	NUM
ajis02-348	278	27	0.17	0.17	NUM
ajis02-348	278	28	0.23	0.23	NUM
ajis02-348	278	29	0.20	0.20	NUM
ajis02-348	278	30	g	g	NOUN
ajis02-348	278	31	,	,	PUNCT
ajis02-348	278	32	0.50	0.50	NUM
ajis02-348	278	33	0.46	0.46	NUM
ajis02-348	278	34	0.47	0.47	NUM
ajis02-348	278	35	pred(0.25	pred(0.25	NOUN
ajis02-348	278	36	)	)	PUNCT
ajis02-348	278	37	0.60	0.60	NUM
ajis02-348	278	38	0.60	0.60	NUM
ajis02-348	278	39	0.60	0.60	NUM
ajis02-348	278	40	0.32	0.32	NUM
ajis02-348	278	41	0.32	0.32	NUM
ajis02-348	278	42	0	0	NUM
ajis02-348	278	43	32	32	NUM
ajis02-348	278	44	wmq	wmq	NOUN
ajis02-348	278	45	0	0	NUM
ajis02-348	278	46	0	0	NUM
ajis02-348	278	47	36	36	NUM
ajis02-348	278	48	36	36	NUM
ajis02-348	278	49	0.36	0.36	NUM
ajis02-348	278	50	mre	mre	PROPN
ajis02-348	278	51	0.5328	0.5328	NUM
ajis02-348	278	52	0.5361	0.5361	NUM
ajis02-348	278	53	0.5352	0.5352	NUM
ajis02-348	278	54	r	r	NOUN
ajis02-348	278	55	sdr	sdr	PROPN
ajis02-348	278	56	0.58	0.58	NUM
ajis02-348	278	57	0.57	0.57	NUM
ajis02-348	278	58	0.57	0.57	NUM
ajis02-348	278	59	for	for	ADP
ajis02-348	278	60	criterion	criterion	NOUN
ajis02-348	278	61	1	1	NUM
ajis02-348	278	62	in	in	ADP
ajis02-348	278	63	table	table	NOUN
ajis02-348	278	64	4	4	NUM
ajis02-348	278	65	,	,	PUNCT
ajis02-348	278	66	the	the	DET
ajis02-348	278	67	sdr	sdr	PROPN
ajis02-348	278	68	is	be	AUX
ajis02-348	278	69	0.57	0.57	NUM
ajis02-348	278	70	(	(	PUNCT
ajis02-348	278	71	the	the	DET
ajis02-348	278	72	largest	large	ADJ
ajis02-348	278	73	sdr	sdr	NOUN
ajis02-348	278	74	in	in	ADP
ajis02-348	278	75	this	this	DET
ajis02-348	278	76	table	table	NOUN
ajis02-348	278	77	)	)	PUNCT
ajis02-348	278	78	,	,	PUNCT
ajis02-348	278	79	so	so	CCONJ
ajis02-348	278	80	the	the	DET
ajis02-348	278	81	estimation	estimation	NOUN
ajis02-348	278	82	is	be	AUX
ajis02-348	278	83	the	the	DET
ajis02-348	278	84	least	least	ADV
ajis02-348	278	85	consistent	consistent	ADJ
ajis02-348	278	86	.	.	PUNCT
ajis02-348	279	1	criterion	criterion	NOUN
ajis02-348	279	2	1	1	NUM
ajis02-348	279	3	in	in	ADP
ajis02-348	279	4	table	table	NOUN
ajis02-348	279	5	5	5	NUM
ajis02-348	279	6	shows	show	VERB
ajis02-348	279	7	that	that	SCONJ
ajis02-348	279	8	the	the	DET
ajis02-348	279	9	calibration	calibration	NOUN
ajis02-348	279	10	does	do	AUX
ajis02-348	279	11	not	not	PART
ajis02-348	279	12	improve	improve	VERB
ajis02-348	279	13	the	the	DET
ajis02-348	279	14	accuracy	accuracy	NOUN
ajis02-348	279	15	.	.	PUNCT
ajis02-348	280	1	for	for	ADP
ajis02-348	280	2	criterion	criterion	NOUN
ajis02-348	280	3	3	3	NUM
ajis02-348	280	4	in	in	ADP
ajis02-348	280	5	table	table	NOUN
ajis02-348	280	6	4	4	NUM
ajis02-348	280	7	,	,	PUNCT
ajis02-348	280	8	the	the	DET
ajis02-348	280	9	sdr	sdr	PROPN
ajis02-348	280	10	is	be	AUX
ajis02-348	280	11	0.53	0.53	NUM
ajis02-348	280	12	.	.	PUNCT
ajis02-348	281	1	this	this	DET
ajis02-348	281	2	consistency	consistency	NOUN
ajis02-348	281	3	is	be	AUX
ajis02-348	281	4	better	well	ADJ
ajis02-348	281	5	than	than	ADP
ajis02-348	281	6	that	that	PRON
ajis02-348	281	7	of	of	ADP
ajis02-348	281	8	criterion	criterion	NOUN
ajis02-348	281	9	1	1	NUM
ajis02-348	281	10	.	.	PUNCT
ajis02-348	282	1	the	the	DET
ajis02-348	282	2	calibration	calibration	NOUN
ajis02-348	282	3	improves	improve	VERB
ajis02-348	282	4	the	the	DET
ajis02-348	282	5	wmq	wmq	NOUN
ajis02-348	282	6	from	from	ADP
ajis02-348	282	7	0.34	0.34	NUM
ajis02-348	282	8	in	in	ADP
ajis02-348	282	9	table	table	NOUN
ajis02-348	282	10	4	4	NUM
ajis02-348	282	11	to	to	PART
ajis02-348	282	12	0.32	0.32	NUM
ajis02-348	282	13	in	in	ADP
ajis02-348	282	14	table	table	NOUN
ajis02-348	282	15	5	5	NUM
ajis02-348	282	16	,	,	PUNCT
ajis02-348	282	17	while	while	SCONJ
ajis02-348	282	18	mre	mre	PROPN
ajis02-348	282	19	remains	remain	VERB
ajis02-348	282	20	unchanged	unchanged	ADJ
ajis02-348	282	21	.	.	PUNCT
ajis02-348	283	1	for	for	ADP
ajis02-348	283	2	criterion	criterion	NOUN
ajis02-348	283	3	2	2	NUM
ajis02-348	283	4	in	in	ADP
ajis02-348	283	5	table	table	NOUN
ajis02-348	283	6	4	4	NUM
ajis02-348	283	7	,	,	PUNCT
ajis02-348	283	8	the	the	DET
ajis02-348	283	9	sdr	sdr	PROPN
ajis02-348	283	10	is	be	AUX
ajis02-348	283	11	0.42	0.42	NUM
ajis02-348	283	12	.	.	PUNCT
ajis02-348	284	1	this	this	DET
ajis02-348	284	2	consistency	consistency	NOUN
ajis02-348	284	3	is	be	AUX
ajis02-348	284	4	much	much	ADV
ajis02-348	284	5	better	well	ADJ
ajis02-348	284	6	than	than	ADP
ajis02-348	284	7	that	that	PRON
ajis02-348	284	8	of	of	ADP
ajis02-348	284	9	criterion	criterion	NOUN
ajis02-348	284	10	1	1	NUM
ajis02-348	284	11	or	or	CCONJ
ajis02-348	284	12	criterion	criterion	NOUN
ajis02-348	284	13	3	3	X
ajis02-348	284	14	.	.	PUNCT
ajis02-348	285	1	the	the	DET
ajis02-348	285	2	calibration	calibration	NOUN
ajis02-348	285	3	improves	improve	VERB
ajis02-348	285	4	the	the	DET
ajis02-348	285	5	wmq	wmq	NOUN
ajis02-348	285	6	from	from	ADP
ajis02-348	285	7	0.44	0.44	NUM
ajis02-348	285	8	in	in	ADP
ajis02-348	285	9	table	table	NOUN
ajis02-348	285	10	4	4	NUM
ajis02-348	285	11	to	to	PART
ajis02-348	285	12	0.32	0.32	NUM
ajis02-348	285	13	in	in	ADP
ajis02-348	285	14	table	table	NOUN
ajis02-348	285	15	5	5	NUM
ajis02-348	285	16	,	,	PUNCT
ajis02-348	285	17	and	and	CCONJ
ajis02-348	285	18	the	the	DET
ajis02-348	285	19	mre	mre	NOUN
ajis02-348	285	20	from	from	ADP
ajis02-348	285	21	0.42	0.42	NUM
ajis02-348	285	22	to	to	ADP
ajis02-348	285	23	0.36	0.36	NUM
ajis02-348	285	24	.	.	PUNCT
ajis02-348	286	1	the	the	DET
ajis02-348	286	2	results	result	NOUN
ajis02-348	286	3	indicate	indicate	VERB
ajis02-348	286	4	that	that	SCONJ
ajis02-348	286	5	the	the	PRON
ajis02-348	286	6	smaller	small	ADJ
ajis02-348	286	7	the	the	DET
ajis02-348	286	8	sdr	sdr	PROPN
ajis02-348	286	9	is	be	AUX
ajis02-348	286	10	,	,	PUNCT
ajis02-348	286	11	the	the	PRON
ajis02-348	286	12	easier	easy	ADJ
ajis02-348	286	13	it	it	PRON
ajis02-348	286	14	is	be	AUX
ajis02-348	286	15	to	to	PART
ajis02-348	286	16	improve	improve	VERB
ajis02-348	286	17	accuracy	accuracy	NOUN
ajis02-348	286	18	by	by	ADP
ajis02-348	286	19	calibration	calibration	NOUN
ajis02-348	286	20	.	.	PUNCT
ajis02-348	287	1	the	the	DET
ajis02-348	287	2	sdr	sdr	PROPN
ajis02-348	287	3	values	value	NOUN
ajis02-348	287	4	remain	remain	VERB
ajis02-348	287	5	the	the	DET
ajis02-348	287	6	same	same	ADJ
ajis02-348	287	7	before	before	ADV
ajis02-348	287	8	and	and	CCONJ
ajis02-348	287	9	after	after	ADP
ajis02-348	287	10	the	the	DET
ajis02-348	287	11	calibration	calibration	NOUN
ajis02-348	287	12	.	.	PUNCT
ajis02-348	288	1	the	the	DET
ajis02-348	288	2	linear	linear	PROPN
ajis02-348	288	3	relationship	relationship	NOUN
ajis02-348	288	4	between	between	ADP
ajis02-348	288	5	estimated	estimate	VERB
ajis02-348	288	6	effort	effort	NOUN
ajis02-348	288	7	and	and	CCONJ
ajis02-348	288	8	actual	actual	ADJ
ajis02-348	288	9	effort	effort	NOUN
ajis02-348	288	10	does	do	AUX
ajis02-348	288	11	not	not	PART
ajis02-348	288	12	change	change	VERB
ajis02-348	288	13	after	after	ADP
ajis02-348	288	14	calibration	calibration	NOUN
ajis02-348	288	15	,	,	PUNCT
ajis02-348	288	16	because	because	SCONJ
ajis02-348	288	17	the	the	DET
ajis02-348	288	18	calibration	calibration	NOUN
ajis02-348	288	19	multiplies	multiply	VERB
ajis02-348	288	20	the	the	DET
ajis02-348	288	21	uncalibrated	uncalibrated	ADJ
ajis02-348	288	22	estimated	estimate	VERB
ajis02-348	288	23	effort	effort	NOUN
ajis02-348	288	24	by	by	ADP
ajis02-348	288	25	a	a	DET
ajis02-348	288	26	constant	constant	ADJ
ajis02-348	288	27	,	,	PUNCT
ajis02-348	288	28	.	.	PUNCT
ajis02-348	289	1	on	on	ADP
ajis02-348	289	2	the	the	DET
ajis02-348	289	3	other	other	ADJ
ajis02-348	289	4	hand	hand	NOUN
ajis02-348	289	5	,	,	PUNCT
ajis02-348	289	6	all	all	DET
ajis02-348	289	7	the	the	DET
ajis02-348	289	8	sdr	sdr	PROPN
ajis02-348	289	9	values	value	NOUN
ajis02-348	289	10	after	after	SCONJ
ajis02-348	289	11	calibration	calibration	NOUN
ajis02-348	289	12	differ	differ	VERB
ajis02-348	289	13	from	from	ADP
ajis02-348	289	14	those	those	PRON
ajis02-348	289	15	before	before	ADP
ajis02-348	289	16	calibration	calibration	NOUN
ajis02-348	289	17	.	.	PUNCT
ajis02-348	290	1	this	this	PRON
ajis02-348	290	2	indicates	indicate	VERB
ajis02-348	290	3	that	that	SCONJ
ajis02-348	290	4	sdr	sdr	PROPN
ajis02-348	290	5	and	and	CCONJ
ajis02-348	290	6	sdr	sdr	PROPN
ajis02-348	290	7	are	be	AUX
ajis02-348	290	8	not	not	PART
ajis02-348	290	9	the	the	DET
ajis02-348	290	10	same	same	ADJ
ajis02-348	290	11	.	.	PUNCT
ajis02-348	291	1	the	the	DET
ajis02-348	291	2	sdr	sdr	PROPN
ajis02-348	291	3	measures	measure	VERB
ajis02-348	291	4	the	the	DET
ajis02-348	291	5	linear	linear	ADJ
ajis02-348	291	6	relationship	relationship	NOUN
ajis02-348	291	7	between	between	ADP
ajis02-348	291	8	estimated	estimate	VERB
ajis02-348	291	9	value	value	NOUN
ajis02-348	291	10	and	and	CCONJ
ajis02-348	291	11	actual	actual	ADJ
ajis02-348	291	12	value	value	NOUN
ajis02-348	291	13	while	while	SCONJ
ajis02-348	291	14	sdr	sdr	PROPN
ajis02-348	291	15	measures	measure	VERB
ajis02-348	291	16	the	the	DET
ajis02-348	291	17	variation	variation	NOUN
ajis02-348	291	18	of	of	ADP
ajis02-348	291	19	relative	relative	ADJ
ajis02-348	291	20	errors	error	NOUN
ajis02-348	291	21	.	.	PUNCT
ajis02-348	292	1	the	the	DET
ajis02-348	292	2	accuracy	accuracy	NOUN
ajis02-348	292	3	results	result	VERB
ajis02-348	292	4	after	after	ADP
ajis02-348	292	5	calibration	calibration	NOUN
ajis02-348	292	6	are	be	AUX
ajis02-348	292	7	comparable	comparable	ADJ
ajis02-348	292	8	with	with	ADP
ajis02-348	292	9	those	those	PRON
ajis02-348	292	10	obtained	obtain	VERB
ajis02-348	292	11	when	when	SCONJ
ajis02-348	292	12	the	the	DET
ajis02-348	292	13	original	original	ADJ
ajis02-348	292	14	formulas	formula	NOUN
ajis02-348	292	15	are	be	AUX
ajis02-348	292	16	used	use	VERB
ajis02-348	292	17	with	with	ADP
ajis02-348	292	18	desharnais	desharnais	PROPN
ajis02-348	292	19	'	'	PART
ajis02-348	292	20	data	datum	NOUN
ajis02-348	292	21	set	set	VERB
ajis02-348	292	22	.	.	PUNCT
ajis02-348	293	1	this	this	DET
ajis02-348	293	2	observation	observation	NOUN
ajis02-348	293	3	further	far	ADV
ajis02-348	293	4	emphasises	emphasise	VERB
ajis02-348	293	5	that	that	SCONJ
ajis02-348	293	6	a	a	DET
ajis02-348	293	7	model	model	NOUN
ajis02-348	293	8	needs	need	VERB
ajis02-348	293	9	to	to	PART
ajis02-348	293	10	be	be	AUX
ajis02-348	293	11	calibrated	calibrate	VERB
ajis02-348	293	12	if	if	SCONJ
ajis02-348	293	13	it	it	PRON
ajis02-348	293	14	is	be	AUX
ajis02-348	293	15	to	to	PART
ajis02-348	293	16	be	be	AUX
ajis02-348	293	17	used	use	VERB
ajis02-348	293	18	in	in	ADP
ajis02-348	293	19	a	a	DET
ajis02-348	293	20	different	different	ADJ
ajis02-348	293	21	environment	environment	NOUN
ajis02-348	293	22	.	.	PUNCT
ajis02-348	294	1	ajis	ajis	ADV
ajis02-348	294	2	vol	vol	VERB
ajis02-348	294	3	5	5	NUM
ajis02-348	294	4	no	no	DET
ajis02-348	294	5	1	1	NUM
ajis02-348	294	6	september	september	NOUN
ajis02-348	294	7	1997	1997	NUM
ajis02-348	294	8	43	43	NUM
ajis02-348	294	9	from	from	ADP
ajis02-348	294	10	table	table	NOUN
ajis02-348	294	11	4	4	NUM
ajis02-348	294	12	and	and	CCONJ
ajis02-348	294	13	table	table	NOUN
ajis02-348	294	14	5	5	NUM
ajis02-348	294	15	it	it	PRON
ajis02-348	294	16	can	can	AUX
ajis02-348	294	17	also	also	ADV
ajis02-348	294	18	be	be	AUX
ajis02-348	294	19	observed	observe	VERB
ajis02-348	294	20	that	that	SCONJ
ajis02-348	294	21	mres	mre	NOUN
ajis02-348	294	22	are	be	AUX
ajis02-348	294	23	greater	great	ADJ
ajis02-348	294	24	than	than	ADP
ajis02-348	294	25	wmqs	wmqs	ADJ
ajis02-348	294	26	in	in	ADP
ajis02-348	294	27	all	all	DET
ajis02-348	294	28	the	the	DET
ajis02-348	294	29	cases	case	NOUN
ajis02-348	294	30	except	except	SCONJ
ajis02-348	294	31	in	in	ADP
ajis02-348	294	32	criterion	criterion	NOUN
ajis02-348	294	33	2	2	NUM
ajis02-348	294	34	of	of	ADP
ajis02-348	294	35	table	table	NOUN
ajis02-348	294	36	4	4	NUM
ajis02-348	294	37	.	.	PUNCT
ajis02-348	294	38	as	as	SCONJ
ajis02-348	294	39	expected	expect	VERB
ajis02-348	294	40	,	,	PUNCT
ajis02-348	294	41	there	there	PRON
ajis02-348	294	42	are	be	VERB
ajis02-348	294	43	extreme	extreme	ADJ
ajis02-348	294	44	values	value	NOUN
ajis02-348	294	45	in	in	ADP
ajis02-348	294	46	cases	case	NOUN
ajis02-348	294	47	where	where	SCONJ
ajis02-348	294	48	mre	mre	PROPN
ajis02-348	294	49	>	>	X
ajis02-348	294	50	wmq	wmq	PROPN
ajis02-348	294	51	,	,	PUNCT
ajis02-348	294	52	and	and	CCONJ
ajis02-348	294	53	the	the	DET
ajis02-348	294	54	formula	formula	NOUN
ajis02-348	294	55	tends	tend	VERB
ajis02-348	294	56	to	to	PART
ajis02-348	294	57	underestimate	underestimate	VERB
ajis02-348	294	58	the	the	DET
ajis02-348	294	59	effort	effort	NOUN
ajis02-348	294	60	in	in	ADP
ajis02-348	294	61	criterion	criterion	NOUN
ajis02-348	294	62	2	2	NUM
ajis02-348	294	63	of	of	ADP
ajis02-348	294	64	table	table	NOUN
ajis02-348	294	65	4	4	NUM
ajis02-348	294	66	,	,	PUNCT
ajis02-348	294	67	where	where	SCONJ
ajis02-348	294	68	mre	mre	PROPN
ajis02-348	294	69	<	<	X
ajis02-348	294	70	wmq	wmq	PROPN
ajis02-348	294	71	.	.	PUNCT
ajis02-348	295	1	summary	summary	NOUN
ajis02-348	295	2	in	in	ADP
ajis02-348	295	3	order	order	NOUN
ajis02-348	295	4	to	to	PART
ajis02-348	295	5	evaluate	evaluate	VERB
ajis02-348	295	6	estimation	estimation	NOUN
ajis02-348	295	7	models	model	NOUN
ajis02-348	295	8	and	and	CCONJ
ajis02-348	295	9	improve	improve	VERB
ajis02-348	295	10	the	the	DET
ajis02-348	295	11	modelling	modelling	NOUN
ajis02-348	295	12	process	process	NOUN
ajis02-348	295	13	,	,	PUNCT
ajis02-348	295	14	we	we	PRON
ajis02-348	295	15	need	need	VERB
ajis02-348	295	16	appropriate	appropriate	ADJ
ajis02-348	295	17	methods	method	NOUN
ajis02-348	295	18	to	to	PART
ajis02-348	295	19	measure	measure	VERB
ajis02-348	295	20	these	these	DET
ajis02-348	295	21	models	model	NOUN
ajis02-348	295	22	and	and	CCONJ
ajis02-348	295	23	determine	determine	VERB
ajis02-348	295	24	model	model	NOUN
ajis02-348	295	25	parameters	parameter	NOUN
ajis02-348	295	26	.	.	PUNCT
ajis02-348	296	1	this	this	DET
ajis02-348	296	2	paper	paper	NOUN
ajis02-348	296	3	reviewed	review	VERB
ajis02-348	296	4	existing	exist	VERB
ajis02-348	296	5	methods	method	NOUN
ajis02-348	296	6	and	and	CCONJ
ajis02-348	296	7	proposed	propose	VERB
ajis02-348	296	8	new	new	ADJ
ajis02-348	296	9	methods	method	NOUN
ajis02-348	296	10	to	to	PART
ajis02-348	296	11	measure	measure	VERB
ajis02-348	296	12	estimation	estimation	NOUN
ajis02-348	296	13	accuracy	accuracy	NOUN
ajis02-348	296	14	and	and	CCONJ
ajis02-348	296	15	consistency	consistency	NOUN
ajis02-348	296	16	,	,	PUNCT
ajis02-348	296	17	and	and	CCONJ
ajis02-348	296	18	to	to	PART
ajis02-348	296	19	determine	determine	VERB
ajis02-348	296	20	model	model	NOUN
ajis02-348	296	21	parameters	parameter	NOUN
ajis02-348	296	22	.	.	PUNCT
ajis02-348	297	1	the	the	DET
ajis02-348	297	2	difference	difference	NOUN
ajis02-348	297	3	measures	measure	NOUN
ajis02-348	297	4	of	of	ADP
ajis02-348	297	5	accuracy	accuracy	NOUN
ajis02-348	297	6	favour	favour	VERB
ajis02-348	297	7	the	the	DET
ajis02-348	297	8	estimation	estimation	NOUN
ajis02-348	297	9	for	for	ADP
ajis02-348	297	10	small	small	ADJ
ajis02-348	297	11	projects	project	NOUN
ajis02-348	297	12	.	.	PUNCT
ajis02-348	298	1	therefore	therefore	ADV
ajis02-348	298	2	it	it	PRON
ajis02-348	298	3	is	be	AUX
ajis02-348	298	4	argued	argue	VERB
ajis02-348	298	5	in	in	ADP
ajis02-348	298	6	this	this	DET
ajis02-348	298	7	paper	paper	NOUN
ajis02-348	298	8	that	that	SCONJ
ajis02-348	298	9	measures	measure	NOUN
ajis02-348	298	10	of	of	ADP
ajis02-348	298	11	accuracy	accuracy	NOUN
ajis02-348	298	12	should	should	AUX
ajis02-348	298	13	be	be	AUX
ajis02-348	298	14	based	base	VERB
ajis02-348	298	15	on	on	ADP
ajis02-348	298	16	relative	relative	ADJ
ajis02-348	298	17	error	error	NOUN
ajis02-348	298	18	of	of	ADP
ajis02-348	298	19	estimation	estimation	NOUN
ajis02-348	298	20	.	.	PUNCT
ajis02-348	299	1	as	as	SCONJ
ajis02-348	299	2	the	the	DET
ajis02-348	299	3	mre	mre	PROPN
ajis02-348	299	4	(	(	PUNCT
ajis02-348	299	5	the	the	DET
ajis02-348	299	6	most	most	ADV
ajis02-348	299	7	widely	widely	ADV
ajis02-348	299	8	used	use	VERB
ajis02-348	299	9	measure	measure	NOUN
ajis02-348	299	10	of	of	ADP
ajis02-348	299	11	accuracy	accuracy	NOUN
ajis02-348	299	12	)	)	PUNCT
ajis02-348	299	13	is	be	AUX
ajis02-348	299	14	influenced	influence	VERB
ajis02-348	299	15	by	by	ADP
ajis02-348	299	16	outliers	outlier	NOUN
ajis02-348	299	17	and	and	CCONJ
ajis02-348	299	18	favours	favour	VERB
ajis02-348	299	19	underestimation	underestimation	NOUN
ajis02-348	299	20	,	,	PUNCT
ajis02-348	299	21	and	and	CCONJ
ajis02-348	299	22	single	single	ADJ
ajis02-348	299	23	value	value	NOUN
ajis02-348	299	24	measures	measure	NOUN
ajis02-348	299	25	are	be	AUX
ajis02-348	299	26	stochastic	stochastic	ADJ
ajis02-348	299	27	,	,	PUNCT
ajis02-348	299	28	the	the	DET
ajis02-348	299	29	wmq	wmq	NOUN
ajis02-348	299	30	is	be	AUX
ajis02-348	299	31	proposed	propose	VERB
ajis02-348	299	32	for	for	ADP
ajis02-348	299	33	accuracy	accuracy	NOUN
ajis02-348	299	34	evaluation	evaluation	NOUN
ajis02-348	299	35	.	.	PUNCT
ajis02-348	300	1	the	the	DET
ajis02-348	300	2	wmq	wmq	NOUN
ajis02-348	300	3	includes	include	VERB
ajis02-348	300	4	more	more	ADJ
ajis02-348	300	5	information	information	NOUN
ajis02-348	300	6	on	on	ADP
ajis02-348	300	7	the	the	DET
ajis02-348	300	8	estimation	estimation	NOUN
ajis02-348	300	9	than	than	ADP
ajis02-348	300	10	single	single	ADJ
ajis02-348	300	11	value	value	NOUN
ajis02-348	300	12	measures	measure	NOUN
ajis02-348	300	13	,	,	PUNCT
ajis02-348	300	14	so	so	SCONJ
ajis02-348	300	15	it	it	PRON
ajis02-348	300	16	is	be	AUX
ajis02-348	300	17	less	less	ADV
ajis02-348	300	18	stochastic	stochastic	ADJ
ajis02-348	300	19	.	.	PUNCT
ajis02-348	301	1	it	it	PRON
ajis02-348	301	2	is	be	AUX
ajis02-348	301	3	also	also	ADV
ajis02-348	301	4	consistent	consistent	ADJ
ajis02-348	301	5	with	with	ADP
ajis02-348	301	6	mre	mre	PROPN
ajis02-348	301	7	when	when	SCONJ
ajis02-348	301	8	there	there	PRON
ajis02-348	301	9	are	be	VERB
ajis02-348	301	10	no	no	DET
ajis02-348	301	11	outliers	outlier	NOUN
ajis02-348	301	12	and	and	CCONJ
ajis02-348	301	13	estimation	estimation	NOUN
ajis02-348	301	14	is	be	AUX
ajis02-348	301	15	unbiased	unbiased	ADJ
ajis02-348	301	16	.	.	PUNCT
ajis02-348	302	1	consistency	consistency	NOUN
ajis02-348	302	2	examines	examine	VERB
ajis02-348	302	3	the	the	DET
ajis02-348	302	4	model	model	NOUN
ajis02-348	302	5	's	's	PART
ajis02-348	302	6	degree	degree	NOUN
ajis02-348	302	7	of	of	ADP
ajis02-348	302	8	ease	ease	NOUN
ajis02-348	302	9	of	of	ADP
ajis02-348	302	10	calibration	calibration	NOUN
ajis02-348	302	11	.	.	PUNCT
ajis02-348	303	1	a	a	DET
ajis02-348	303	2	consistently	consistently	ADV
ajis02-348	303	3	overestimating	overestimate	VERB
ajis02-348	303	4	or	or	CCONJ
ajis02-348	303	5	underestimating	underestimating	ADJ
ajis02-348	303	6	model	model	NOUN
ajis02-348	303	7	is	be	AUX
ajis02-348	303	8	more	more	ADV
ajis02-348	303	9	easily	easily	ADV
ajis02-348	303	10	calibrated	calibrate	VERB
ajis02-348	303	11	than	than	ADP
ajis02-348	303	12	an	an	DET
ajis02-348	303	13	inconsistent	inconsistent	ADJ
ajis02-348	303	14	one	one	NUM
ajis02-348	303	15	.	.	PUNCT
ajis02-348	304	1	the	the	DET
ajis02-348	304	2	correlation	correlation	NOUN
ajis02-348	304	3	coefficient	coefficient	NOUN
ajis02-348	304	4	r	r	NOUN
ajis02-348	304	5	between	between	ADP
ajis02-348	304	6	observed	observed	ADJ
ajis02-348	304	7	and	and	CCONJ
ajis02-348	304	8	actual	actual	ADJ
ajis02-348	304	9	values	value	NOUN
ajis02-348	304	10	has	have	AUX
ajis02-348	304	11	been	be	AUX
ajis02-348	304	12	used	use	VERB
ajis02-348	304	13	to	to	PART
ajis02-348	304	14	evaluate	evaluate	VERB
ajis02-348	304	15	consistency	consistency	NOUN
ajis02-348	304	16	.	.	PUNCT
ajis02-348	305	1	the	the	DET
ajis02-348	305	2	r	r	NOUN
ajis02-348	305	3	favours	favour	VERB
ajis02-348	305	4	a	a	DET
ajis02-348	305	5	data	data	NOUN
ajis02-348	305	6	set	set	VERB
ajis02-348	305	7	with	with	ADP
ajis02-348	305	8	large	large	ADJ
ajis02-348	305	9	variance	variance	NOUN
ajis02-348	305	10	.	.	PUNCT
ajis02-348	306	1	it	it	PRON
ajis02-348	306	2	is	be	AUX
ajis02-348	306	3	determined	determine	VERB
ajis02-348	306	4	not	not	PART
ajis02-348	306	5	only	only	ADV
ajis02-348	306	6	by	by	ADP
ajis02-348	306	7	estimation	estimation	NOUN
ajis02-348	306	8	but	but	CCONJ
ajis02-348	306	9	also	also	ADV
ajis02-348	306	10	by	by	ADP
ajis02-348	306	11	the	the	DET
ajis02-348	306	12	distribution	distribution	NOUN
ajis02-348	306	13	of	of	ADP
ajis02-348	306	14	the	the	DET
ajis02-348	306	15	actual	actual	ADJ
ajis02-348	306	16	values	value	NOUN
ajis02-348	306	17	.	.	PUNCT
ajis02-348	307	1	in	in	ADP
ajis02-348	307	2	this	this	DET
ajis02-348	307	3	paper	paper	NOUN
ajis02-348	307	4	,	,	PUNCT
ajis02-348	307	5	the	the	DET
ajis02-348	307	6	standard	standard	ADJ
ajis02-348	307	7	deviation	deviation	NOUN
ajis02-348	307	8	of	of	ADP
ajis02-348	307	9	the	the	DET
ajis02-348	307	10	ratios	ratio	NOUN
ajis02-348	307	11	of	of	ADP
ajis02-348	307	12	the	the	DET
ajis02-348	307	13	estimate	estimate	NOUN
ajis02-348	307	14	to	to	ADP
ajis02-348	307	15	actual	actual	ADJ
ajis02-348	307	16	effort	effort	NOUN
ajis02-348	307	17	(	(	PUNCT
ajis02-348	307	18	sdr	sdr	NOUN
ajis02-348	307	19	)	)	PUNCT
ajis02-348	307	20	is	be	AUX
ajis02-348	307	21	proposed	propose	VERB
ajis02-348	307	22	as	as	ADP
ajis02-348	307	23	a	a	DET
ajis02-348	307	24	measure	measure	NOUN
ajis02-348	307	25	of	of	ADP
ajis02-348	307	26	consistency	consistency	NOUN
ajis02-348	307	27	.	.	PUNCT
ajis02-348	308	1	the	the	DET
ajis02-348	308	2	sdr	sdr	PROPN
ajis02-348	308	3	is	be	AUX
ajis02-348	308	4	a	a	DET
ajis02-348	308	5	measure	measure	NOUN
ajis02-348	308	6	of	of	ADP
ajis02-348	308	7	the	the	DET
ajis02-348	308	8	variation	variation	NOUN
ajis02-348	308	9	or	or	CCONJ
ajis02-348	308	10	spread	spread	NOUN
ajis02-348	308	11	of	of	ADP
ajis02-348	308	12	the	the	DET
ajis02-348	308	13	relative	relative	ADJ
ajis02-348	308	14	error	error	NOUN
ajis02-348	308	15	.	.	PUNCT
ajis02-348	309	1	the	the	DET
ajis02-348	309	2	method	method	NOUN
ajis02-348	309	3	of	of	ADP
ajis02-348	309	4	least	least	ADJ
ajis02-348	309	5	squares	square	NOUN
ajis02-348	309	6	is	be	AUX
ajis02-348	309	7	the	the	DET
ajis02-348	309	8	conventional	conventional	ADJ
ajis02-348	309	9	regression	regression	NOUN
ajis02-348	309	10	criterion	criterion	NOUN
ajis02-348	309	11	,	,	PUNCT
ajis02-348	309	12	but	but	CCONJ
ajis02-348	309	13	it	it	PRON
ajis02-348	309	14	is	be	AUX
ajis02-348	309	15	based	base	VERB
ajis02-348	309	16	on	on	ADP
ajis02-348	309	17	a	a	DET
ajis02-348	309	18	difference	difference	NOUN
ajis02-348	309	19	measure	measure	NOUN
ajis02-348	309	20	,	,	PUNCT
ajis02-348	309	21	which	which	PRON
ajis02-348	309	22	is	be	AUX
ajis02-348	309	23	not	not	PART
ajis02-348	309	24	suitable	suitable	ADJ
ajis02-348	309	25	for	for	ADP
ajis02-348	309	26	software	software	NOUN
ajis02-348	309	27	cost	cost	NOUN
ajis02-348	309	28	estimation	estimation	NOUN
ajis02-348	309	29	.	.	PUNCT
ajis02-348	310	1	the	the	DET
ajis02-348	310	2	criterion	criterion	NOUN
ajis02-348	310	3	of	of	ADP
ajis02-348	310	4	minimising	minimise	VERB
ajis02-348	310	5	the	the	DET
ajis02-348	310	6	mre	mre	NOUN
ajis02-348	310	7	will	will	AUX
ajis02-348	310	8	produce	produce	VERB
ajis02-348	310	9	a	a	DET
ajis02-348	310	10	formula	formula	NOUN
ajis02-348	310	11	which	which	PRON
ajis02-348	310	12	tends	tend	VERB
ajis02-348	310	13	to	to	PART
ajis02-348	310	14	underestimate	underestimate	VERB
ajis02-348	310	15	the	the	DET
ajis02-348	310	16	effort	effort	NOUN
ajis02-348	310	17	.	.	PUNCT
ajis02-348	311	1	to	to	PART
ajis02-348	311	2	overcome	overcome	VERB
ajis02-348	311	3	the	the	DET
ajis02-348	311	4	shortcomings	shortcoming	NOUN
ajis02-348	311	5	of	of	ADP
ajis02-348	311	6	these	these	DET
ajis02-348	311	7	criteria	criterion	NOUN
ajis02-348	311	8	,	,	PUNCT
ajis02-348	311	9	this	this	DET
ajis02-348	311	10	paper	paper	NOUN
ajis02-348	311	11	proposes	propose	VERB
ajis02-348	311	12	to	to	PART
ajis02-348	311	13	use	use	VERB
ajis02-348	311	14	the	the	DET
ajis02-348	311	15	criterion	criterion	NOUN
ajis02-348	311	16	of	of	ADP
ajis02-348	311	17	least	least	ADJ
ajis02-348	311	18	squares	square	NOUN
ajis02-348	311	19	of	of	ADP
ajis02-348	311	20	the	the	DET
ajis02-348	311	21	logarithmic	logarithmic	ADJ
ajis02-348	311	22	ratio	ratio	NOUN
ajis02-348	311	23	of	of	ADP
ajis02-348	311	24	estimate	estimate	NOUN
ajis02-348	311	25	to	to	ADP
ajis02-348	311	26	actual	actual	ADJ
ajis02-348	311	27	value	value	NOUN
ajis02-348	311	28	for	for	ADP
ajis02-348	311	29	regression	regression	NOUN
ajis02-348	311	30	.	.	PUNCT
ajis02-348	312	1	the	the	DET
ajis02-348	312	2	proposed	propose	VERB
ajis02-348	312	3	measures	measure	NOUN
ajis02-348	312	4	and	and	CCONJ
ajis02-348	312	5	criterion	criterion	NOUN
ajis02-348	312	6	share	share	VERB
ajis02-348	312	7	the	the	DET
ajis02-348	312	8	characteristic	characteristic	NOUN
ajis02-348	312	9	that	that	SCONJ
ajis02-348	312	10	they	they	PRON
ajis02-348	312	11	are	be	AUX
ajis02-348	312	12	based	base	VERB
ajis02-348	312	13	on	on	ADP
ajis02-348	312	14	relative	relative	ADJ
ajis02-348	312	15	errors	error	NOUN
ajis02-348	312	16	of	of	ADP
ajis02-348	312	17	the	the	DET
ajis02-348	312	18	estimation	estimation	NOUN
ajis02-348	312	19	.	.	PUNCT
ajis02-348	313	1	applied	apply	VERB
ajis02-348	313	2	to	to	ADP
ajis02-348	313	3	real	real	ADJ
ajis02-348	313	4	-	-	PUNCT
ajis02-348	313	5	world	world	NOUN
ajis02-348	313	6	data	datum	NOUN
ajis02-348	313	7	,	,	PUNCT
ajis02-348	313	8	the	the	DET
ajis02-348	313	9	proposed	propose	VERB
ajis02-348	313	10	measures	measure	NOUN
ajis02-348	313	11	and	and	CCONJ
ajis02-348	313	12	criterion	criterion	NOUN
ajis02-348	313	13	appear	appear	VERB
ajis02-348	313	14	superior	superior	ADJ
ajis02-348	313	15	to	to	ADP
ajis02-348	313	16	the	the	DET
ajis02-348	313	17	measures	measure	NOUN
ajis02-348	313	18	and	and	CCONJ
ajis02-348	313	19	criteria	criterion	NOUN
ajis02-348	313	20	that	that	PRON
ajis02-348	313	21	are	be	AUX
ajis02-348	313	22	currently	currently	ADV
ajis02-348	313	23	used	use	VERB
ajis02-348	313	24	in	in	ADP
ajis02-348	313	25	the	the	DET
ajis02-348	313	26	field	field	NOUN
ajis02-348	313	27	and	and	CCONJ
ajis02-348	313	28	reported	report	VERB
ajis02-348	313	29	in	in	ADP
ajis02-348	313	30	the	the	DET
ajis02-348	313	31	literature	literature	NOUN
ajis02-348	313	32	.	.	PUNCT
ajis02-348	314	1	it	it	PRON
ajis02-348	314	2	is	be	AUX
ajis02-348	314	3	therefore	therefore	ADV
ajis02-348	314	4	recommended	recommend	VERB
ajis02-348	314	5	that	that	SCONJ
ajis02-348	314	6	the	the	DET
ajis02-348	314	7	proposed	propose	VERB
ajis02-348	314	8	measures	measure	NOUN
ajis02-348	314	9	and	and	CCONJ
ajis02-348	314	10	criterion	criterion	NOUN
ajis02-348	314	11	be	be	AUX
ajis02-348	314	12	used	use	VERB
ajis02-348	314	13	for	for	ADP
ajis02-348	314	14	further	further	ADJ
ajis02-348	314	15	assessment	assessment	NOUN
ajis02-348	314	16	references	reference	NOUN
ajis02-348	314	17	abdel	abdel	PROPN
ajis02-348	314	18	-	-	PUNCT
ajis02-348	314	19	hamid	hamid	PROPN
ajis02-348	314	20	,	,	PUNCT
ajis02-348	314	21	t.k	t.k	PROPN
ajis02-348	314	22	.	.	PROPN
ajis02-348	314	23	(	(	PUNCT
ajis02-348	314	24	1990	1990	NUM
ajis02-348	314	25	)	)	PUNCT
ajis02-348	314	26	"	"	PUNCT
ajis02-348	314	27	on	on	ADP
ajis02-348	314	28	the	the	DET
ajis02-348	314	29	utility	utility	NOUN
ajis02-348	314	30	of	of	ADP
ajis02-348	314	31	historical	historical	ADJ
ajis02-348	314	32	project	project	NOUN
ajis02-348	314	33	statistics	statistic	NOUN
ajis02-348	314	34	for	for	ADP
ajis02-348	314	35	cost	cost	NOUN
ajis02-348	314	36	and	and	CCONJ
ajis02-348	314	37	schedule	schedule	NOUN
ajis02-348	314	38	estimation	estimation	NOUN
ajis02-348	314	39	:	:	PUNCT
ajis02-348	314	40	results	result	NOUN
ajis02-348	314	41	from	from	ADP
ajis02-348	314	42	a	a	DET
ajis02-348	314	43	simulation	simulation	NOUN
ajis02-348	314	44	-	-	PUNCT
ajis02-348	314	45	based	base	VERB
ajis02-348	314	46	case	case	NOUN
ajis02-348	314	47	study	study	NOUN
ajis02-348	314	48	"	"	PUNCT
ajis02-348	314	49	,	,	PUNCT
ajis02-348	314	50	journal	journal	NOUN
ajis02-348	314	51	of	of	ADP
ajis02-348	314	52	systems	system	NOUN
ajis02-348	314	53	&	&	CCONJ
ajis02-348	314	54	software	software	PROPN
ajis02-348	314	55	,	,	PUNCT
ajis02-348	314	56	sep	sep	PROPN
ajis02-348	314	57	.	.	PROPN
ajis02-348	314	58	,	,	PUNCT
ajis02-348	314	59	vol.13	vol.13	NOUN
ajis02-348	314	60	,	,	PUNCT
ajis02-348	314	61	no.l	no.l	NOUN
ajis02-348	314	62	,	,	PUNCT
ajis02-348	314	63	pp.71	pp.71	NOUN
ajis02-348	314	64	-	-	PUNCT
ajis02-348	314	65	82	82	NUM
ajis02-348	314	66	.	.	PUNCT
ajis02-348	315	1	albrecht	albrecht	PROPN
ajis02-348	315	2	,	,	PUNCT
ajis02-348	315	3	aj	aj	PROPN
ajis02-348	315	4	.	.	PROPN
ajis02-348	315	5	&	&	CCONJ
ajis02-348	315	6	gaffney	gaffney	PROPN
ajis02-348	315	7	,	,	PUNCT
ajis02-348	315	8	j.	j.	PROPN
ajis02-348	315	9	(	(	PUNCT
ajis02-348	315	10	1983	1983	NUM
ajis02-348	315	11	)	)	PUNCT
ajis02-348	315	12	"	"	PUNCT
ajis02-348	315	13	software	software	NOUN
ajis02-348	315	14	function	function	NOUN
ajis02-348	315	15	,	,	PUNCT
ajis02-348	315	16	source	source	NOUN
ajis02-348	315	17	lines	line	NOUN
ajis02-348	315	18	of	of	ADP
ajis02-348	315	19	code	code	NOUN
ajis02-348	315	20	,	,	PUNCT
ajis02-348	315	21	and	and	CCONJ
ajis02-348	315	22	development	development	NOUN
ajis02-348	315	23	effort	effort	NOUN
ajis02-348	315	24	prediction	prediction	NOUN
ajis02-348	315	25	:	:	PUNCT
ajis02-348	315	26	a	a	DET
ajis02-348	315	27	software	software	NOUN
ajis02-348	315	28	science	science	NOUN
ajis02-348	315	29	validation	validation	NOUN
ajis02-348	315	30	"	"	PUNCT
ajis02-348	315	31	,	,	PUNCT
ajis02-348	315	32	ieee	ieee	NOUN
ajis02-348	315	33	transactions	transaction	NOUN
ajis02-348	315	34	on	on	ADP
ajis02-348	315	35	software	software	NOUN
ajis02-348	315	36	engineering	engineering	NOUN
ajis02-348	315	37	,	,	PUNCT
ajis02-348	315	38	nov	nov	PROPN
ajis02-348	315	39	.	.	PROPN
ajis02-348	315	40	,	,	PUNCT
ajis02-348	315	41	vol.9	vol.9	PROPN
ajis02-348	315	42	,	,	PUNCT
ajis02-348	315	43	no.6	no.6	PROPN
ajis02-348	315	44	,	,	PUNCT
ajis02-348	315	45	pp.639	pp.639	PROPN
ajis02-348	315	46	-	-	PUNCT
ajis02-348	315	47	648	648	NUM
ajis02-348	315	48	.	.	PUNCT
ajis02-348	316	1	boehm	boehm	PROPN
ajis02-348	316	2	,	,	PUNCT
ajis02-348	316	3	b.w.(1981	b.w.(1981	PROPN
ajis02-348	316	4	)	)	PUNCT
ajis02-348	316	5	software	software	NOUN
ajis02-348	316	6	engineering	engineering	NOUN
ajis02-348	316	7	economics	economic	NOUN
ajis02-348	316	8	,	,	PUNCT
ajis02-348	316	9	prentice	prentice	NOUN
ajis02-348	316	10	-	-	PUNCT
ajis02-348	316	11	hall	hall	NOUN
ajis02-348	316	12	,	,	PUNCT
ajis02-348	316	13	inc	inc	PROPN
ajis02-348	316	14	.	.	PROPN
ajis02-348	316	15	,	,	PUNCT
ajis02-348	316	16	englewood	englewood	PROPN
ajis02-348	316	17	cliffs	cliffs	PROPN
ajis02-348	316	18	,	,	PUNCT
ajis02-348	316	19	new	new	PROPN
ajis02-348	316	20	jersey	jersey	PROPN
ajis02-348	316	21	,	,	PUNCT
ajis02-348	316	22	.	.	PUNCT
ajis02-348	317	1	conte	conte	PROPN
ajis02-348	317	2	,	,	PUNCT
ajis02-348	317	3	s.d	s.d	PROPN
ajis02-348	317	4	.	.	PROPN
ajis02-348	317	5	,	,	PUNCT
ajis02-348	317	6	dunsmore	dunsmore	PROPN
ajis02-348	317	7	,	,	PUNCT
ajis02-348	317	8	h.e	h.e	PROPN
ajis02-348	317	9	.	.	PROPN
ajis02-348	317	10	&	&	CCONJ
ajis02-348	317	11	shen	shen	PROPN
ajis02-348	317	12	,	,	PUNCT
ajis02-348	317	13	v.y	v.y	PROPN
ajis02-348	317	14	.	.	PROPN
ajis02-348	317	15	(	(	PUNCT
ajis02-348	317	16	1986	1986	NUM
ajis02-348	317	17	)	)	PUNCT
ajis02-348	317	18	software	software	NOUN
ajis02-348	317	19	engineering	engineering	NOUN
ajis02-348	317	20	metrics	metric	NOUN
ajis02-348	317	21	and	and	CCONJ
ajis02-348	317	22	models	model	NOUN
ajis02-348	317	23	,	,	PUNCT
ajis02-348	317	24	benjamin	benjamin	PROPN
ajis02-348	317	25	/	/	SYM
ajis02-348	317	26	cummings	cummings	PROPN
ajis02-348	317	27	publishing	publish	VERB
ajis02-348	317	28	company	company	PROPN
ajis02-348	317	29	,	,	PUNCT
ajis02-348	317	30	inc	inc	PROPN
ajis02-348	317	31	.	.	PROPN
ajis02-348	317	32	,	,	PUNCT
ajis02-348	317	33	menlo	menlo	PROPN
ajis02-348	317	34	park	park	PROPN
ajis02-348	317	35	,	,	PUNCT
ajis02-348	317	36	calif	calif	PROPN
ajis02-348	317	37	..	..	PROPN
ajis02-348	317	38	deeson	deeson	PROPN
ajis02-348	317	39	,	,	PUNCT
ajis02-348	317	40	e.	e.	PROPN
ajis02-348	317	41	(	(	PUNCT
ajis02-348	317	42	1991	1991	NUM
ajis02-348	317	43	)	)	PUNCT
ajis02-348	317	44	collins	collins	PROPN
ajis02-348	317	45	dictionary	dictionary	NOUN
ajis02-348	317	46	of	of	ADP
ajis02-348	317	47	information	information	NOUN
ajis02-348	317	48	technology	technology	NOUN
ajis02-348	317	49	,	,	PUNCT
ajis02-348	317	50	harper	harper	PROPN
ajis02-348	317	51	collins	collins	PROPN
ajis02-348	317	52	publishers	publisher	NOUN
ajis02-348	317	53	,	,	PUNCT
ajis02-348	317	54	glasgow	glasgow	PROPN
ajis02-348	317	55	.	.	PUNCT
ajis02-348	318	1	desharnais	desharnais	PROPN
ajis02-348	318	2	,	,	PUNCT
ajis02-348	318	3	j	j	PROPN
ajis02-348	318	4	-	-	PUNCT
ajis02-348	318	5	m.	m.	NOUN
ajis02-348	318	6	(	(	PUNCT
ajis02-348	318	7	1988	1988	NUM
ajis02-348	318	8	)	)	PUNCT
ajis02-348	318	9	"	"	PUNCT
ajis02-348	318	10	programme	programme	NOUN
ajis02-348	318	11	de	de	X
ajis02-348	318	12	maitrise	maitrise	X
ajis02-348	318	13	en	en	X
ajis02-348	318	14	informatique	informatique	PROPN
ajis02-348	318	15	de	de	X
ajis02-348	318	16	gestion	gestion	PROPN
ajis02-348	318	17	"	"	PUNCT
ajis02-348	318	18	,	,	PUNCT
ajis02-348	318	19	university	university	PROPN
ajis02-348	318	20	of	of	ADP
ajis02-348	318	21	montreal	montreal	PROPN
ajis02-348	318	22	,	,	PUNCT
ajis02-348	318	23	canada	canada	PROPN
ajis02-348	318	24	.	.	PUNCT
ajis02-348	318	25	fenton	fenton	PROPN
ajis02-348	318	26	,	,	PUNCT
ajis02-348	318	27	n.	n.	NOUN
ajis02-348	318	28	(	(	PUNCT
ajis02-348	318	29	1994	1994	NUM
ajis02-348	318	30	)	)	PUNCT
ajis02-348	318	31	"	"	PUNCT
ajis02-348	318	32	software	software	NOUN
ajis02-348	318	33	measurement	measurement	NOUN
ajis02-348	318	34	:	:	PUNCT
ajis02-348	318	35	a	a	DET
ajis02-348	318	36	necessary	necessary	ADJ
ajis02-348	318	37	scientific	scientific	ADJ
ajis02-348	318	38	basis	basis	NOUN
ajis02-348	318	39	"	"	PUNCT
ajis02-348	318	40	,	,	PUNCT
ajis02-348	318	41	ieee	ieee	NOUN
ajis02-348	318	42	transactions	transaction	NOUN
ajis02-348	318	43	on	on	ADP
ajis02-348	318	44	software	software	NOUN
ajis02-348	318	45	engineering	engineering	NOUN
ajis02-348	318	46	,	,	PUNCT
ajis02-348	318	47	mar	mar	PROPN
ajis02-348	318	48	.	.	PROPN
ajis02-348	318	49	,	,	PUNCT
ajis02-348	318	50	vol.20	vol.20	PROPN
ajis02-348	318	51	,	,	PUNCT
ajis02-348	318	52	no.3	no.3	VERB
ajis02-348	318	53	,	,	PUNCT
ajis02-348	318	54	pp	pp	PROPN
ajis02-348	318	55	.	.	PUNCT
ajis02-348	319	1	199	199	NUM
ajis02-348	319	2	-	-	SYM
ajis02-348	319	3	206	206	NUM
ajis02-348	319	4	.	.	PUNCT
ajis02-348	320	1	gao	gao	PROPN
ajis02-348	320	2	,	,	PUNCT
ajis02-348	320	3	x.	x.	PROPN
ajis02-348	320	4	&	&	CCONJ
ajis02-348	320	5	lo	lo	PROPN
ajis02-348	320	6	,	,	PUNCT
ajis02-348	320	7	b.	b.	PROPN
ajis02-348	320	8	(	(	PUNCT
ajis02-348	320	9	1995	1995	NUM
ajis02-348	320	10	)	)	PUNCT
ajis02-348	320	11	"	"	PUNCT
ajis02-348	320	12	an	an	DET
ajis02-348	320	13	integrated	integrated	ADJ
ajis02-348	320	14	software	software	NOUN
ajis02-348	320	15	cost	cost	NOUN
ajis02-348	320	16	model	model	NOUN
ajis02-348	320	17	based	base	VERB
ajis02-348	320	18	on	on	ADP
ajis02-348	320	19	cocomo	cocomo	NOUN
ajis02-348	320	20	and	and	CCONJ
ajis02-348	320	21	function	function	NOUN
ajis02-348	320	22	point	point	NOUN
ajis02-348	320	23	analysis	analysis	NOUN
ajis02-348	320	24	"	"	PUNCT
ajis02-348	320	25	,	,	PUNCT
ajis02-348	320	26	proceedings	proceeding	NOUN
ajis02-348	320	27	of	of	ADP
ajis02-348	320	28	the	the	DET
ajis02-348	320	29	software	software	NOUN
ajis02-348	320	30	education	education	NOUN
ajis02-348	320	31	conference	conference	NOUN
ajis02-348	320	32	(	(	PUNCT
ajis02-348	320	33	srig	srig	NOUN
ajis02-348	320	34	-	-	PUNCT
ajis02-348	320	35	et94	et94	PROPN
ajis02-348	320	36	)	)	PUNCT
ajis02-348	320	37	,	,	PUNCT
ajis02-348	320	38	pp.86	pp.86	NOUN
ajis02-348	320	39	-	-	PUNCT
ajis02-348	320	40	93	93	NUM
ajis02-348	320	41	,	,	PUNCT
ajis02-348	320	42	ieee	ieee	NOUN
ajis02-348	320	43	computer	computer	NOUN
ajis02-348	320	44	society	society	NOUN
ajis02-348	320	45	,	,	PUNCT
ajis02-348	320	46	los	los	PROPN
ajis02-348	320	47	alamitos	alamitos	PROPN
ajis02-348	320	48	,	,	PUNCT
ajis02-348	320	49	ca	ca	PROPN
ajis02-348	320	50	.	.	PUNCT
ajis02-348	321	1	jenson	jenson	PROPN
ajis02-348	321	2	,	,	PUNCT
ajis02-348	321	3	r.l	r.l	PROPN
ajis02-348	321	4	.	.	PROPN
ajis02-348	321	5	&	&	CCONJ
ajis02-348	321	6	bartley	bartley	PROPN
ajis02-348	321	7	,	,	PUNCT
ajis02-348	321	8	j.w	j.w	PROPN
ajis02-348	321	9	.	.	PROPN
ajis02-348	321	10	(	(	PUNCT
ajis02-348	321	11	1991	1991	NUM
ajis02-348	321	12	)	)	PUNCT
ajis02-348	321	13	"	"	PUNCT
ajis02-348	321	14	parametric	parametric	ADJ
ajis02-348	321	15	estimation	estimation	NOUN
ajis02-348	321	16	of	of	ADP
ajis02-348	321	17	programming	programming	NOUN
ajis02-348	321	18	effort	effort	NOUN
ajis02-348	321	19	:	:	PUNCT
ajis02-348	321	20	an	an	DET
ajis02-348	321	21	object	object	NOUN
ajis02-348	321	22	-	-	PUNCT
ajis02-348	321	23	oriented	orient	VERB
ajis02-348	321	24	model	model	NOUN
ajis02-348	321	25	"	"	PUNCT
ajis02-348	321	26	,	,	PUNCT
ajis02-348	321	27	journal	journal	NOUN
ajis02-348	321	28	of	of	ADP
ajis02-348	321	29	systems	system	NOUN
ajis02-348	321	30	&	&	CCONJ
ajis02-348	321	31	software	software	PROPN
ajis02-348	321	32	,	,	PUNCT
ajis02-348	321	33	may	may	AUX
ajis02-348	321	34	,	,	PUNCT
ajis02-348	321	35	vol.15	vol.15	NOUN
ajis02-348	321	36	,	,	PUNCT
ajis02-348	321	37	no.2	no.2	PROPN
ajis02-348	321	38	,	,	PUNCT
ajis02-348	321	39	pp.107	pp.107	PROPN
ajis02-348	321	40	-	-	SYM
ajis02-348	321	41	114	114	NUM
ajis02-348	321	42	.	.	PUNCT
ajis02-348	322	1	j0rgensen	j0rgensen	PROPN
ajis02-348	322	2	,	,	PUNCT
ajis02-348	322	3	m.	m.	NOUN
ajis02-348	322	4	(	(	PUNCT
ajis02-348	322	5	1994	1994	NUM
ajis02-348	322	6	)	)	PUNCT
ajis02-348	322	7	"	"	PUNCT
ajis02-348	322	8	a	a	DET
ajis02-348	322	9	comparative	comparative	ADJ
ajis02-348	322	10	study	study	NOUN
ajis02-348	322	11	of	of	ADP
ajis02-348	322	12	software	software	NOUN
ajis02-348	322	13	maintenance	maintenance	NOUN
ajis02-348	322	14	effort	effort	NOUN
ajis02-348	322	15	prediction	prediction	NOUN
ajis02-348	322	16	models	model	NOUN
ajis02-348	322	17	"	"	PUNCT
ajis02-348	322	18	,	,	PUNCT
ajis02-348	322	19	proceedings	proceeding	NOUN
ajis02-348	322	20	of	of	ADP
ajis02-348	322	21	5th	5th	ADJ
ajis02-348	322	22	australian	australian	ADJ
ajis02-348	322	23	conference	conference	NOUN
ajis02-348	322	24	on	on	ADP
ajis02-348	322	25	information	information	NOUN
ajis02-348	322	26	systems	system	NOUN
ajis02-348	322	27	,	,	PUNCT
ajis02-348	322	28	27	27	NUM
ajis02-348	322	29	-	-	SYM
ajis02-348	322	30	29	29	NUM
ajis02-348	322	31	september	september	PROPN
ajis02-348	322	32	1994	1994	NUM
ajis02-348	322	33	,	,	PUNCT
ajis02-348	322	34	melbourne	melbourne	PROPN
ajis02-348	322	35	,	,	PUNCT
ajis02-348	322	36	australia	australia	PROPN
ajis02-348	322	37	,	,	PUNCT
ajis02-348	322	38	pp.699	pp.699	PROPN
ajis02-348	322	39	-	-	PUNCT
ajis02-348	322	40	709	709	NUM
ajis02-348	322	41	.	.	PUNCT
ajis02-348	323	1	kan	kan	PROPN
ajis02-348	323	2	,	,	PUNCT
ajis02-348	323	3	s.h	s.h	PROPN
ajis02-348	323	4	.	.	PROPN
ajis02-348	323	5	,	,	PUNCT
ajis02-348	323	6	basili	basili	NOUN
ajis02-348	323	7	,	,	PUNCT
ajis02-348	323	8	v.r	v.r	PROPN
ajis02-348	323	9	.	.	PROPN
ajis02-348	323	10	&	&	CCONJ
ajis02-348	323	11	shapiro	shapiro	PROPN
ajis02-348	323	12	,	,	PUNCT
ajis02-348	323	13	l.n	l.n	PROPN
ajis02-348	323	14	.	.	PROPN
ajis02-348	323	15	(	(	PUNCT
ajis02-348	323	16	1994	1994	NUM
ajis02-348	323	17	)	)	PUNCT
ajis02-348	323	18	"	"	PUNCT
ajis02-348	323	19	software	software	NOUN
ajis02-348	323	20	quality	quality	NOUN
ajis02-348	323	21	:	:	PUNCT
ajis02-348	323	22	an	an	DET
ajis02-348	323	23	overview	overview	NOUN
ajis02-348	323	24	from	from	ADP
ajis02-348	323	25	the	the	DET
ajis02-348	323	26	perspective	perspective	NOUN
ajis02-348	323	27	of	of	ADP
ajis02-348	323	28	total	total	ADJ
ajis02-348	323	29	quality	quality	NOUN
ajis02-348	323	30	management	management	NOUN
ajis02-348	323	31	"	"	PUNCT
ajis02-348	323	32	,	,	PUNCT
ajis02-348	323	33	ibm	ibm	PROPN
ajis02-348	323	34	systems	system	NOUN
ajis02-348	323	35	journal	journal	PROPN
ajis02-348	323	36	,	,	PUNCT
ajis02-348	323	37	mar	mar	PROPN
ajis02-348	323	38	,	,	PUNCT
ajis02-348	323	39	vol.33	vol.33	PROPN
ajis02-348	323	40	,	,	PUNCT
ajis02-348	323	41	no.l	no.l	NOUN
ajis02-348	323	42	,	,	PUNCT
ajis02-348	323	43	pp.4	pp.4	PROPN
ajis02-348	323	44	-	-	PUNCT
ajis02-348	323	45	19	19	NUM
ajis02-348	323	46	.	.	PUNCT
ajis02-348	323	47	kenkel	kenkel	PROPN
ajis02-348	323	48	,	,	PUNCT
ajis02-348	323	49	j.l	j.l	PROPN
ajis02-348	323	50	.	.	PROPN
ajis02-348	323	51	(	(	PUNCT
ajis02-348	323	52	1989	1989	NUM
ajis02-348	323	53	)	)	PUNCT
ajis02-348	323	54	introductory	introductory	ADJ
ajis02-348	323	55	statistics	statistic	NOUN
ajis02-348	323	56	for	for	ADP
ajis02-348	323	57	management	management	NOUN
ajis02-348	323	58	and	and	CCONJ
ajis02-348	323	59	economics	economic	NOUN
ajis02-348	323	60	(	(	PUNCT
ajis02-348	323	61	third	third	ADJ
ajis02-348	323	62	edition	edition	NOUN
ajis02-348	323	63	)	)	PUNCT
ajis02-348	323	64	,	,	PUNCT
ajis02-348	323	65	duxbury	duxbury	PROPN
ajis02-348	323	66	press	press	PROPN
ajis02-348	323	67	,	,	PUNCT
ajis02-348	323	68	usa	usa	PROPN
ajis02-348	323	69	.	.	PROPN
ajis02-348	323	70	khoshgoftaar	khoshgoftaar	PROPN
ajis02-348	323	71	,	,	PUNCT
ajis02-348	323	72	t.m	t.m	PROPN
ajis02-348	323	73	.	.	PROPN
ajis02-348	323	74	,	,	PUNCT
ajis02-348	323	75	j.c	j.c	PROPN
ajis02-348	323	76	.	.	PROPN
ajis02-348	323	77	munson	munson	PROPN
ajis02-348	323	78	,	,	PUNCT
ajis02-348	323	79	et	et	PROPN
ajis02-348	323	80	al	al	PROPN
ajis02-348	323	81	(	(	PUNCT
ajis02-348	323	82	1992	1992	NUM
ajis02-348	323	83	)	)	PUNCT
ajis02-348	323	84	"	"	PUNCT
ajis02-348	323	85	predictive	predictive	ADJ
ajis02-348	323	86	modeling	modeling	NOUN
ajis02-348	323	87	techniques	technique	NOUN
ajis02-348	323	88	of	of	ADP
ajis02-348	323	89	software	software	NOUN
ajis02-348	323	90	quality	quality	NOUN
ajis02-348	323	91	from	from	ADP
ajis02-348	323	92	software	software	NOUN
ajis02-348	323	93	measures	measure	NOUN
ajis02-348	323	94	"	"	PUNCT
ajis02-348	323	95	,	,	PUNCT
ajis02-348	323	96	ieee	ieee	NOUN
ajis02-348	323	97	transactions	transaction	NOUN
ajis02-348	323	98	on	on	ADP
ajis02-348	323	99	software	software	NOUN
ajis02-348	323	100	engineering	engineering	NOUN
ajis02-348	323	101	,	,	PUNCT
ajis02-348	323	102	nov	nov	PROPN
ajis02-348	323	103	.	.	PROPN
ajis02-348	323	104	,	,	PUNCT
ajis02-348	323	105	vol.18	vol.18	NOUN
ajis02-348	323	106	,	,	PUNCT
ajis02-348	323	107	no.l	no.l	NOUN
ajis02-348	323	108	1	1	NUM
ajis02-348	323	109	,	,	PUNCT
ajis02-348	323	110	pp.979	pp.979	NOUN
ajis02-348	323	111	-	-	SYM
ajis02-348	323	112	987	987	NUM
ajis02-348	323	113	.	.	PUNCT
ajis02-348	324	1	ajis	ajis	ADV
ajis02-348	324	2	vol	vol	VERB
ajis02-348	324	3	5	5	NUM
ajis02-348	324	4	no	no	DET
ajis02-348	324	5	1	1	NUM
ajis02-348	324	6	september	september	NOUN
ajis02-348	324	7	1997	1997	NUM
ajis02-348	324	8	44	44	NUM
ajis02-348	324	9	kitchenham	kitchenham	PROPN
ajis02-348	324	10	,	,	PUNCT
ajis02-348	324	11	b.a	b.a	PROPN
ajis02-348	324	12	.	.	PROPN
ajis02-348	324	13	,	,	PUNCT
ajis02-348	324	14	kok	kok	PROPN
ajis02-348	324	15	,	,	PUNCT
ajis02-348	324	16	p.a.m	p.a.m	PROPN
ajis02-348	324	17	.	.	PROPN
ajis02-348	324	18	&	&	CCONJ
ajis02-348	324	19	kirakowski	kirakowski	PROPN
ajis02-348	324	20	,	,	PUNCT
ajis02-348	324	21	j.	j.	PROPN
ajis02-348	324	22	(	(	PUNCT
ajis02-348	324	23	1990	1990	NUM
ajis02-348	324	24	)	)	PUNCT
ajis02-348	324	25	"	"	PUNCT
ajis02-348	324	26	the	the	DET
ajis02-348	324	27	mermaid	mermaid	NOUN
ajis02-348	324	28	approach	approach	NOUN
ajis02-348	324	29	to	to	ADP
ajis02-348	324	30	software	software	NOUN
ajis02-348	324	31	cost	cost	NOUN
ajis02-348	324	32	estimation	estimation	NOUN
ajis02-348	324	33	"	"	PUNCT
ajis02-348	324	34	,	,	PUNCT
ajis02-348	324	35	proceedings	proceeding	NOUN
ajis02-348	324	36	of	of	ADP
ajis02-348	324	37	the	the	DET
ajis02-348	324	38	annual	annual	ADJ
ajis02-348	324	39	esprit	esprit	NOUN
ajis02-348	324	40	conference	conference	NOUN
ajis02-348	324	41	,	,	PUNCT
ajis02-348	324	42	brussels	brussels	PROPN
ajis02-348	324	43	,	,	PUNCT
ajis02-348	324	44	november	november	PROPN
ajis02-348	324	45	12	12	NUM
ajis02-348	324	46	-	-	SYM
ajis02-348	324	47	15	15	NUM
ajis02-348	324	48	,	,	PUNCT
ajis02-348	324	49	1990	1990	NUM
ajis02-348	324	50	,	,	PUNCT
ajis02-348	324	51	pp.296	pp.296	NOUN
ajis02-348	324	52	-	-	SYM
ajis02-348	324	53	314	314	NUM
ajis02-348	324	54	.	.	PUNCT
ajis02-348	325	1	lederer	lederer	PROPN
ajis02-348	325	2	,	,	PUNCT
ajis02-348	325	3	a.l	a.l	PROPN
ajis02-348	325	4	.	.	PROPN
ajis02-348	325	5	&	&	CCONJ
ajis02-348	325	6	prasad	prasad	PROPN
ajis02-348	325	7	,	,	PUNCT
ajis02-348	325	8	j.	j.	PROPN
ajis02-348	325	9	(	(	PUNCT
ajis02-348	325	10	1993	1993	NUM
ajis02-348	325	11	)	)	PUNCT
ajis02-348	325	12	"	"	PUNCT
ajis02-348	325	13	information	information	NOUN
ajis02-348	325	14	systems	system	NOUN
ajis02-348	325	15	software	software	NOUN
ajis02-348	325	16	cost	cost	NOUN
ajis02-348	325	17	estimating	estimate	VERB
ajis02-348	325	18	:	:	PUNCT
ajis02-348	325	19	a	a	DET
ajis02-348	325	20	current	current	ADJ
ajis02-348	325	21	assessment	assessment	NOUN
ajis02-348	325	22	"	"	PUNCT
ajis02-348	325	23	,	,	PUNCT
ajis02-348	325	24	journal	journal	NOUN
ajis02-348	325	25	of	of	ADP
ajis02-348	325	26	information	information	NOUN
ajis02-348	325	27	technology	technology	NOUN
ajis02-348	325	28	,	,	PUNCT
ajis02-348	325	29	no.8	no.8	PROPN
ajis02-348	325	30	,	,	PUNCT
ajis02-348	325	31	pp.22	pp.22	NOUN
ajis02-348	325	32	-	-	PUNCT
ajis02-348	325	33	33	33	NUM
ajis02-348	325	34	.	.	PUNCT
ajis02-348	326	1	miyazaki	miyazaki	PROPN
ajis02-348	326	2	,	,	PUNCT
ajis02-348	326	3	y.	y.	PROPN
ajis02-348	326	4	takanou	takanou	PROPN
ajis02-348	326	5	,	,	PUNCT
ajis02-348	326	6	a.	a.	PROPN
ajis02-348	326	7	&	&	CCONJ
ajis02-348	326	8	nozaki	nozaki	PROPN
ajis02-348	326	9	,	,	PUNCT
ajis02-348	326	10	h.	h.	PROPN
ajis02-348	326	11	(	(	PUNCT
ajis02-348	326	12	1991	1991	NUM
ajis02-348	326	13	)	)	PUNCT
ajis02-348	326	14	"	"	PUNCT
ajis02-348	326	15	method	method	NOUN
ajis02-348	326	16	to	to	PART
ajis02-348	326	17	estimate	estimate	VERB
ajis02-348	326	18	parameter	parameter	NOUN
ajis02-348	326	19	values	value	NOUN
ajis02-348	326	20	in	in	ADP
ajis02-348	326	21	software	software	NOUN
ajis02-348	326	22	prediction	prediction	NOUN
ajis02-348	326	23	models	model	NOUN
ajis02-348	326	24	"	"	PUNCT
ajis02-348	326	25	,	,	PUNCT
ajis02-348	326	26	information	information	NOUN
ajis02-348	326	27	&	&	CCONJ
ajis02-348	326	28	software	software	PROPN
ajis02-348	326	29	technology	technology	PROPN
ajis02-348	326	30	,	,	PUNCT
ajis02-348	326	31	apr	apr	PROPN
ajis02-348	326	32	.	.	PROPN
ajis02-348	326	33	,	,	PUNCT
ajis02-348	326	34	vol.33	vol.33	PROPN
ajis02-348	326	35	,	,	PUNCT
ajis02-348	326	36	no.3	no.3	PROPN
ajis02-348	326	37	,	,	PUNCT
ajis02-348	326	38	pp.239	pp.239	PROPN
ajis02-348	326	39	-	-	PUNCT
ajis02-348	326	40	243	243	PROPN
ajis02-348	326	41	.	.	PUNCT
ajis02-348	327	1	mukhopadhyay	mukhopadhyay	NOUN
ajis02-348	327	2	,	,	PUNCT
ajis02-348	327	3	t.	t.	PROPN
ajis02-348	327	4	,	,	PUNCT
ajis02-348	327	5	vicinanza	vicinanza	PROPN
ajis02-348	327	6	,	,	PUNCT
ajis02-348	327	7	s.s	s.s	PROPN
ajis02-348	327	8	.	.	PROPN
ajis02-348	327	9	&	&	CCONJ
ajis02-348	327	10	prietula	prietula	PROPN
ajis02-348	327	11	,	,	PUNCT
ajis02-348	327	12	m.j	m.j	PROPN
ajis02-348	327	13	.	.	PROPN
ajis02-348	327	14	(	(	PUNCT
ajis02-348	327	15	1992	1992	NUM
ajis02-348	327	16	)	)	PUNCT
ajis02-348	327	17	"	"	PUNCT
ajis02-348	327	18	examining	examine	VERB
ajis02-348	327	19	the	the	DET
ajis02-348	327	20	feasibility	feasibility	NOUN
ajis02-348	327	21	of	of	ADP
ajis02-348	327	22	a	a	DET
ajis02-348	327	23	case	case	NOUN
ajis02-348	327	24	-	-	PUNCT
ajis02-348	327	25	based	base	VERB
ajis02-348	327	26	reasoning	reasoning	NOUN
ajis02-348	327	27	model	model	NOUN
ajis02-348	327	28	for	for	ADP
ajis02-348	327	29	software	software	NOUN
ajis02-348	327	30	effort	effort	NOUN
ajis02-348	327	31	estimation	estimation	NOUN
ajis02-348	327	32	"	"	PUNCT
ajis02-348	327	33	,	,	PUNCT
ajis02-348	327	34	mis	mis	PROPN
ajis02-348	327	35	quarterly	quarterly	PROPN
ajis02-348	327	36	,	,	PUNCT
ajis02-348	327	37	jun	jun	PROPN
ajis02-348	327	38	.	.	PROPN
ajis02-348	327	39	,	,	PUNCT
ajis02-348	328	1	pp	pp	PROPN
ajis02-348	328	2	.	.	PUNCT
ajis02-348	329	1	155	155	NUM
ajis02-348	329	2	-	-	SYM
ajis02-348	329	3	170	170	NUM
ajis02-348	329	4	.	.	PUNCT
ajis02-348	330	1	pfleeger	pfleeger	NOUN
ajis02-348	330	2	,	,	PUNCT
ajis02-348	330	3	s.l	s.l	PROPN
ajis02-348	330	4	.	.	PROPN
ajis02-348	330	5	(	(	PUNCT
ajis02-348	330	6	1991	1991	NUM
ajis02-348	330	7	)	)	PUNCT
ajis02-348	330	8	"	"	PUNCT
ajis02-348	330	9	model	model	NOUN
ajis02-348	330	10	of	of	ADP
ajis02-348	330	11	software	software	NOUN
ajis02-348	330	12	effort	effort	NOUN
ajis02-348	330	13	and	and	CCONJ
ajis02-348	330	14	productivity	productivity	NOUN
ajis02-348	330	15	"	"	PUNCT
ajis02-348	330	16	,	,	PUNCT
ajis02-348	330	17	information	information	NOUN
ajis02-348	330	18	&	&	CCONJ
ajis02-348	330	19	software	software	PROPN
ajis02-348	330	20	technology	technology	PROPN
ajis02-348	330	21	,	,	PUNCT
ajis02-348	330	22	apr	apr	PROPN
ajis02-348	330	23	.	.	PROPN
ajis02-348	330	24	,	,	PUNCT
ajis02-348	330	25	vol.33	vol.33	PROPN
ajis02-348	330	26	,	,	PUNCT
ajis02-348	330	27	no.3	no.3	NOUN
ajis02-348	330	28	,	,	PUNCT
ajis02-348	330	29	pp.224	pp.224	NOUN
ajis02-348	330	30	-	-	SYM
ajis02-348	330	31	231	231	NUM
ajis02-348	330	32	.	.	PUNCT
ajis02-348	331	1	rowlands	rowlands	PROPN
ajis02-348	331	2	,	,	PUNCT
ajis02-348	331	3	b.h	b.h	PROPN
ajis02-348	331	4	.	.	PUNCT
ajis02-348	331	5	(	(	PUNCT
ajis02-348	331	6	1989	1989	NUM
ajis02-348	331	7	)	)	PUNCT
ajis02-348	331	8	"	"	PUNCT
ajis02-348	331	9	estimating	estimate	VERB
ajis02-348	331	10	instructional	instructional	ADJ
ajis02-348	331	11	systems	system	NOUN
ajis02-348	331	12	development	development	NOUN
ajis02-348	331	13	effort	effort	NOUN
ajis02-348	331	14	and	and	CCONJ
ajis02-348	331	15	costs	cost	NOUN
ajis02-348	331	16	"	"	PUNCT
ajis02-348	331	17	,	,	PUNCT
ajis02-348	331	18	proceedings	proceeding	NOUN
ajis02-348	331	19	of	of	ADP
ajis02-348	331	20	the	the	DET
ajis02-348	331	21	7th	7th	ADJ
ajis02-348	331	22	annual	annual	ADJ
ajis02-348	331	23	conf	conf	NOUN
ajis02-348	331	24	.	.	PUNCT
ajis02-348	331	25	of	of	ADP
ajis02-348	331	26	the	the	DET
ajis02-348	331	27	australian	australian	ADJ
ajis02-348	331	28	society	society	NOUN
ajis02-348	331	29	for	for	ADP
ajis02-348	331	30	computers	computer	NOUN
ajis02-348	331	31	in	in	ADP
ajis02-348	331	32	learning	learn	VERB
ajis02-348	331	33	in	in	ADP
ajis02-348	331	34	tertiary	tertiary	ADJ
ajis02-348	331	35	education	education	NOUN
ajis02-348	331	36	,	,	PUNCT
ajis02-348	331	37	gold	gold	NOUN
ajis02-348	331	38	coast	coast	NOUN
ajis02-348	331	39	,	,	PUNCT
ajis02-348	331	40	1989	1989	NUM
ajis02-348	331	41	,	,	PUNCT
ajis02-348	331	42	pp.340	pp.340	PROPN
ajis02-348	331	43	-	-	PUNCT
ajis02-348	331	44	348	348	NUM
ajis02-348	331	45	.	.	PUNCT
ajis02-348	332	1	wrigley	wrigley	PROPN
ajis02-348	332	2	,	,	PUNCT
ajis02-348	332	3	c.d	c.d	PROPN
ajis02-348	332	4	.	.	PROPN
ajis02-348	332	5	&	&	CCONJ
ajis02-348	332	6	dexter	dexter	PROPN
ajis02-348	332	7	,	,	PUNCT
ajis02-348	332	8	a.s	a.s	PROPN
ajis02-348	332	9	.	.	PROPN
ajis02-348	332	10	(	(	PUNCT
ajis02-348	332	11	1991	1991	NUM
ajis02-348	332	12	)	)	PUNCT
ajis02-348	332	13	"	"	PUNCT
ajis02-348	332	14	a	a	DET
ajis02-348	332	15	model	model	NOUN
ajis02-348	332	16	for	for	ADP
ajis02-348	332	17	measuring	measure	VERB
ajis02-348	332	18	information	information	NOUN
ajis02-348	332	19	system	system	NOUN
ajis02-348	332	20	size	size	NOUN
ajis02-348	332	21	"	"	PUNCT
ajis02-348	332	22	,	,	PUNCT
ajis02-348	332	23	mis	mis	PROPN
ajis02-348	332	24	quarterly	quarterly	PROPN
ajis02-348	332	25	,	,	PUNCT
ajis02-348	332	26	jun	jun	PROPN
ajis02-348	332	27	.	.	PROPN
ajis02-348	332	28	,	,	PUNCT
ajis02-348	332	29	pp.245	pp.245	PROPN
ajis02-348	332	30	-	-	PUNCT
ajis02-348	332	31	257	257	NUM
ajis02-348	332	32	.	.	PUNCT
ajis02-348	333	1	verner	verner	NOUN
ajis02-348	333	2	,	,	PUNCT
ajis02-348	333	3	j.	j.	PROPN
ajis02-348	333	4	&	&	CCONJ
ajis02-348	333	5	tate	tate	PROPN
ajis02-348	333	6	,	,	PUNCT
ajis02-348	333	7	g.	g.	PROPN
ajis02-348	333	8	(	(	PUNCT
ajis02-348	333	9	1992	1992	NUM
ajis02-348	333	10	)	)	PUNCT
ajis02-348	333	11	"	"	PUNCT
ajis02-348	333	12	a	a	DET
ajis02-348	333	13	software	software	NOUN
ajis02-348	333	14	size	size	NOUN
ajis02-348	333	15	model	model	NOUN
ajis02-348	333	16	"	"	PUNCT
ajis02-348	333	17	,	,	PUNCT
ajis02-348	333	18	ieee	ieee	NOUN
ajis02-348	333	19	transactions	transaction	NOUN
ajis02-348	333	20	on	on	ADP
ajis02-348	333	21	software	software	NOUN
ajis02-348	333	22	engineering	engineering	NOUN
ajis02-348	333	23	,	,	PUNCT
ajis02-348	333	24	apr	apr	PROPN
ajis02-348	333	25	.	.	PROPN
ajis02-348	333	26	,	,	PUNCT
ajis02-348	333	27	vol.l8,no.4,pp.265	vol.l8,no.4,pp.265	VERB
ajis02-348	333	28	-	-	SYM
ajis02-348	333	29	278	278	NUM
ajis02-348	333	30	.	.	PUNCT
ajis02-348	334	1	ajis	ajis	ADV
ajis02-348	334	2	vol	vol	VERB
ajis02-348	334	3	5	5	NUM
ajis02-348	334	4	no	no	DET
ajis02-348	334	5	1	1	NUM
ajis02-348	334	6	september	september	NOUN
ajis02-348	334	7	1997	1997	NUM