| { |
| "paper_id": "J76-4001", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T02:52:05.641122Z" |
| }, |
| "title": "FIFTH INTERNATIONAL SYMPOSIUM ON THE USE OF COMPUTERS I N", |
| "authors": [ |
| { |
| "first": "Internationpl", |
| "middle": [ |
| "C" |
| ], |
| "last": "O M M I T T E E", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "F", |
| "middle": [ |
| "V" |
| ], |
| "last": "Spechtler", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "R" |
| ], |
| "last": "Allen", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Jones", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "I", |
| "middle": [ |
| "T" |
| ], |
| "last": "Piirainen", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "L", |
| "middle": [], |
| "last": "Finland", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "France", |
| "middle": [], |
| "last": "Fossier", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "W", |
| "middle": [], |
| "last": "Lenders", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "M", |
| "middle": [ |
| "L" |
| ], |
| "last": "Alinei", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "C" |
| ], |
| "last": "Holland", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Hong", |
| "middle": [], |
| "last": "Loh", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "F", |
| "middle": [], |
| "last": "Kong", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Papp", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Iceland", |
| "middle": [], |
| "last": "J6nsson", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "K '" |
| ], |
| "last": "Havanur", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "India", |
| "middle": [ |
| "; U" |
| ], |
| "last": "Oman", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "L", |
| "middle": [ |
| "F" |
| ], |
| "last": "Lara", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "K", |
| "middle": [], |
| "last": "Hyldgaard-Jensen", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Joyce", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Raben", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "S" |
| ], |
| "last": "North", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "D", |
| "middle": [ |
| "E" |
| ], |
| "last": "Ager", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "James", |
| "middle": [ |
| "E" |
| ], |
| "last": "George", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Gerald", |
| "middle": [ |
| "L" |
| ], |
| "last": "Engel", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "E", |
| "middle": [ |
| "D" |
| ], |
| "last": "I T E D B Y D A N I E", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "L", |
| "middle": [ |
| "G" |
| ], |
| "last": "Bobrow", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Allan", |
| "middle": [ |
| "C" |
| ], |
| "last": "C~l I N S", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "New", |
| "middle": [], |
| "last": "York", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "", |
| "middle": [ |
| "E F" |
| ], |
| "last": "Lindblom", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Stephen", |
| "middle": [ |
| "F" |
| ], |
| "last": "Weiss A N D", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Donald", |
| "middle": [ |
| "F" |
| ], |
| "last": "S T A N A T", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Hirschbert", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "F" |
| ], |
| "last": "Weiss", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "D", |
| "middle": [ |
| "F" |
| ], |
| "last": "Stanat", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "G", |
| "middle": [], |
| "last": "A -Mago", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "suitable for reproduction in the proceedings. Typed, double spaced, up to 15 pages. 8\"xlO\" pictorial matter, glossy B&W photographs or photographable drawings. Title page. Authors' names, complete mailing add~ess, telephone numbers, if multiple, indicate which handles correspondence and wLll deliver the talk. Each page should have the principal author's name .on it.", |
| "pdf_parse": { |
| "paper_id": "J76-4001", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "suitable for reproduction in the proceedings. Typed, double spaced, up to 15 pages. 8\"xlO\" pictorial matter, glossy B&W photographs or photographable drawings. Title page. Authors' names, complete mailing add~ess, telephone numbers, if multiple, indicate which handles correspondence and wLll deliver the talk. Each page should have the principal author's name .on it.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "There ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "C h a l r r a n of t h e Y e r n b e r s h l p Committee, was p r e p a r f n q e c a r p a l a n t o recruit neb m e m b e r s , P e t r i c k r e v l e~e d tne s t a t u s of t h e A J C L i n t h e a b s e n c e o f p a v e H a y s , i t s W i t o r , In d i s c u s s i n g A a y s * recent s u r v e y of t h e T e r n b e r s h i D a b o u t t h e J a q r n a l , p e t r i c~ remarked O n its q u a l i t y a n d t h b r o u u h n e s s , b a t h in P r e D a r a t i o n a n d in t h e a n a l y s i s o f t h e r e s u l t s , Over 2 0 0 members r e s~o n d e d , a n u n u s u a l l y high p e r c e n t a g e ; t h e y s t r o n q l r~ s u p p o r t e d c o n t i n u l n a p u b l i c a t i o n in microfiche form.", |
| "sec_num": null |
| }, |
| { |
| "text": "A J C L a ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "x~e n s e s a s s o c i a t e c j w i t h t h e", |
| "sec_num": null |
| }, |
| { |
| "text": "N", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I n s t i t u t e of L~n g u i s t i c s i n t h e Acarjemy o f S c i e n c e s o f t h e USSR, o s t e n s i b l y on t h e basis of a l e t t e r h e h a d s u b m i t t e d t o t h e", |
| "sec_num": null |
| }, |
| { |
| "text": "f o r t h e Program C o m m i t t e e .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "L k e r a n n o u n c e d t h a t t h e n e x t I q t e r n a t icnal J o i n t C o n t e r p n c r o n A r t i t $ c i a 1 l n t e l lisence N i l 1 h e h e l d 2 2 -2 b A \\ I P \"~~ 1~ 7 7 a t t h e ~~a s s a c h u s e t t s 1 n s t f t u t e o f T e c h n a 1 o g y in C a~b r i d q e , s 4 a s s a c h u b e t t s , J a n e p o b i~s~n r e p o r t e d o n Local 4 r r a n s e m e n t s r w i t h p d r t lculal' e m a h a s i s a, t h e b a n q u e t s c h e d u l e d f o r t h e e v e n j n u r ~1 3 0 r t l y a f t e r t h e c o n c l u s i o n o f t h e b u s l n e s s M e e t f n a * P a u l c h a~i n ~e~o r t c d", |
| "sec_num": null |
| }, |
| { |
| "text": "assistance.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Of t h e 21 a b s t r a c t s s u b m i t t e d , 1 4 w e r e a c c e p t e d ; h e e x a r e s s e d h i s a p m r e c i a t i o n to t b e C o m a i t t e e m e m n e r s tor t h e i r", |
| "sec_num": null |
| }, |
| { |
| "text": "J o n a t h a n A l l e n r M I T S e c r e t a r y l r e a s u r e r : ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "i s e x p e r i e n c e h i t h p u b l i c i t y a b o u t t h e C a l l f o r P a p e r s s w~e s t e d t h a t a C h e c k l i s t b e e s t a b l i s h e j t o p r o v i d e m o r e e f f e c t j v e n o t i t j c a t i o n , H O D b a r n e s r e p o r t e d f o r t h e f i a m i n a t l o n C o m m i t t e e t h a t t h e f o l l o h i n q s l a t e o f o f ' f i c e r s h a d b e e n p r g p o s e d : f ' a e s i d e n t :-F ' a u l C h a p i n , hSC V i c e P r e s i d e n t :", |
| "sec_num": null |
| }, |
| { |
| "text": "A t h e journal. 1", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "annual r e p o r t to y o u t y p i c a l l y c o n s i s t s o f s t a t e m e n t s a b o u t m e m b e r s h i p a n d finances, Tnis k i l l b e a t y p i c a l r e p o r t , w h e n t h e n e N j o u r n a l uas f i r s t i s s u e d in 1 9 7 4 t h e r e was, a dra-at1.c i n c ? r e a s e in t h e number o f A C L members-from j u s t u n d e r 100 in 1 9 7 3 , to o v e r 800 b y e a r l y 1 9 7 5 , S i n c e the,n, t h e s e i m p r e s s i v e g a i n s h a v e b e e n s o s e r i o u s l y e r o d e d t h a t o u r C u r r e n t m e h b e f s h i p s t a n d s at 5 8 0 ( 4 4 5 i n d i v i d u a l s a n d 1 3 5 institUtlons).", |
| "sec_num": null |
| }, |
| { |
| "text": "a", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T h e r e c e n t l y e s t a b l i s h e d m e t h o d O f b i l l i n g m e m b e r s f o r t h e i r a n n u d l d u e s ( I n c l u d i n g t h e d u e s notice on o n e o f t h e o p a q u e c a r d s i n t h e journal) w h i c h w a s c o n c e i v e d", |
| "sec_num": null |
| }, |
| { |
| "text": "ACLJ S e c r e t a r y -l r e a s u r e r ' s R e P o x t r (c, a c t a b c r 3 P 7 6 P a g e 9", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Mood R o b e r t s", |
| "sec_num": null |
| }, |
| { |
| "text": "14,353.37", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "R e c e i p t s : P e m b e r s h i p d u e s -1 9 7 5 -1 q 7 6", |
| "sec_num": null |
| }, |
| { |
| "text": "L~x s h u r s e m e n t s:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "S l h r 8 1 8 m 6 1", |
| "sec_num": null |
| }, |
| { |
| "text": "3 1 7 S 4~7 7 7 . 6 4 ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "A d n l i n 5 S t r a t l V e c 6 S t s r o f f i c e s u p p l i e s , m d 1 L i n q r a n d A J C L c o a t s n o t c o v e r e d b y C A L account", |
| "sec_num": null |
| }, |
| { |
| "text": "Graphics, Historical studies, Information retrieval; Input techniques, Lexicography, Literary stylistics; Medieval studies; Music; Photocomposition, Public sexvice systems, Sernaptics .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "THEMES Frontiers between language and literature, Fine arts;", |
| "sec_num": null |
| }, |
| { |
| "text": "FOURTH ANNUAL CONFERENCE C O M P U T C R G R A P H I C S A i 4 D I N T E R A C -r I V E . T E C H N I Q U m E S C A L L FOR PAPERS TOP I CS DEADLINE", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "THEMES Frontiers between language and literature, Fine arts;", |
| "sec_num": null |
| }, |
| { |
| "text": "v i d u a l l y , w c i~r r i v c d a t a slightly L .& J P w +-r 'La c L G C k s w .r , L . G f m ... - 5 -I '5 L - 5 G rc. . . . *f f C .- t . C - , , E: * d s C* * yr r , C, c L 0 C C * $a *F F.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "In a t t e m p t i n g t o rcvicw t h c pilpcrs t h a t nppcnr i n this book c o l l c c t i v c l y r a t h e r tllan i n d i", |
| "sec_num": null |
| }, |
| { |
| "text": "1 7 . S c h a n k , R . ''Ilsjng Knowlcdgc t o I J n d e r~t a n d~~ TINLAP rrocccd i n g s pp. 1 1 7 -1 2 1 , J u n e 3 9 7 5 . The book under review contains the proceedings of a small conference ( 2 2 p a r t i c i p a n t s ) w i t h t h e same t i t l e , h e l d i n October ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "In a t t e m p t i n g t o rcvicw t h c pilpcrs t h a t nppcnr i n this book c o l l c c t i v c l y r a t h e r tllan i n d i", |
| "sec_num": null |
| }, |
| { |
| "text": "1973", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "American Journal of Computational Linguistics", |
| "sec_num": null |
| }, |
| { |
| "text": "operations on it is given b e l a w .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "A precise formulation of the polynomial and t h e", |
| "sec_num": null |
| }, |
| { |
| "text": "A semigroup is formally defined as an ordered pair < S , -i where Note that all coef iicients of R(M) arc either 1 or 0. We wi 11 adopt the usual convention of not explicitly writing 1 for the terms with that coefficient and omitting telms with a coefficient of 0.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Algebraic preliminaqies, and notation", |
| "sec_num": "111." |
| }, |
| { |
| "text": "A --context-free grammar is a system G = <VN, VT, P, S> where VN and V are finite, disjoint, non erlpty sets denoted non-terminal and T terminal symbols respectively. We denote by V the set V : I VT.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Algebraic preliminaqies, and notation", |
| "sec_num": "111." |
| }, |
| { |
| "text": "The N symbol S is the distinguished nonterminal from which all derivations b e g i n , and P 2 s the set of productions of G. A context-free grammer is proper if it does not contain productions of thz form A -+ c (erasures) or A B where A and E are both nonterminals.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Algebraic preliminaqies, and notation", |
| "sec_num": "111." |
| }, |
| { |
| "text": "It can easily be shown that the set of Languages generated by", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Algebraic preliminaqies, and notation", |
| "sec_num": "111." |
| }, |
| { |
| "text": "proper context-free grammars is exactly the set of context-free languages. In addition, an arbitrary context-free grammar can be made proper by a straightforward method which alters the structure of the grammar very little. In this study we will deal with only", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Algebraic preliminaqies, and notation", |
| "sec_num": "111." |
| }, |
| { |
| "text": "proper .context-free grammars. This guarantees that all terminal strings have a finite number of derivations i n C-, and thus makes possible our goal of finding all derivations of an input.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Algebraic preliminaqies, and notation", |
| "sec_num": "111." |
| }, |
| { |
| "text": "Productibns of G will be indexed by integers. Thus A M denotes th that A -+ >I is the i production in P. We will deal only with l e f tmost deriyations. A leftmost derivation is completely specifzed by the initial sentential form and the sequence of production indices. If-", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "i", |
| "sec_num": null |
| }, |
| { |
| "text": "A c_ 1 is the sequence of production indices in the leftmost derivation.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "+ + C of N 6 -V from M c V , we rite ?I -N.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "The length of a derivation D is denoted by I , and is equal to the number of production indices in L .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "We will use, but not formally define, the notion of height of a The algebraic structure used in this work is the semiring of polynomials R(H -I*) where H = H (v) I t h e free half -group generated by V, and I is the in'dex set of the set of proJuctions P. We will.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "use an initial segment of the n a t u r a l numbers, Construction for the proof: , 3 ) 1 ( B , 4 ) 1 [ ( B , 4 ) 1", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 81, |
| "end": 112, |
| "text": ", 3 ) 1 ( B , 4 ) 1 [ ( B , 4 )", |
| "ref_id": "FIGREF14" |
| } |
| ], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "Let V = v1 IJ V", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "+ Lemma J: Let M E V , n A E V", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "*", |
| "sec_num": null |
| }, |
| { |
| "text": "Case 2.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "A p p l y i n g 6 w e g e t", |
| "sec_num": null |
| }, |
| { |
| "text": "V1 = V.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "A p p l y i n g 6 w e g e t", |
| "sec_num": null |
| }, |
| { |
| "text": "Example 2:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T h e e n t i r e j o b of parsing is now done by g , since t h e polynomial p i s equal t o ( A , .' I) . Hence t h e p a r s i n g polynomial is", |
| "sec_num": null |
| }, |
| { |
| "text": "We use t h e s a m e grammar and i n p u t s t r i r i g as above. A , B , a , b) . ", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 67, |
| "end": 81, |
| "text": "A , B , a , b)", |
| "ref_id": null |
| } |
| ], |
| "eq_spans": [], |
| "section": "T h e e n t i r e j o b of parsing is now done by g , since t h e polynomial p i s equal t o ( A , .' I) . Hence t h e p a r s i n g polynomial is", |
| "sec_num": null |
| }, |
| { |
| "text": "Automatic indexing may be considered to be a two-step process:. ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Theories of Term Importance", |
| "sec_num": "1." |
| }, |
| { |
| "text": "The conclusion is t h a t t h e b e s t terms a r e those whose document frequency k k B , o r t o t a l frequency F , i~ n e i t h e r too l a r g e nor too small, and whose o f t h e c o i l e c t i o n : c o n t r a r i w i s e , terms with low document f r e q u e n c i e s must be made more g e n e r a l by i n c r e a s i n g their assignment frequencies. [ 5 ] This can be achieved by j o i n i n g two o r more high frequency terms i n t o term p h r a s e s ,", |
| "cite_spans": [ |
| { |
| "start": 364, |
| "end": 369, |
| "text": "[ 5 ]", |
| "ref_id": "BIBREF7" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "while assembling a number of low frequency terms i n t o term classes.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "Obviously, a term phrase exhibits a lower assignment frequency t h a n any phrase where r and hk a r e t h e number of documents c o n t a i n i n g t e r m k t h a t are k r e l e v a n t and n o n r e l e v a n t r e s p e c t i v e l y t o query Q, and I R I and I I I i r e t h e t o t a l number o f r e l e v a n t and n o n r e l e v a n t documents f o r t h a t query.;' When a t e r m k o c c u r s in more t h a n one q u e r y , i t s term r e l e v a n c e may be t a k e n a s t h e t h o s e query t e r m s which are p r e v a l e n t i n t h e r e l e v a n t items and r h e i n t h e nonrelevant, and vice-versa f o r t h a s e previl3cn-t mainly %n t h e n o n r e l e v a n t .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "Furthermore, t h e terms f a l l i n g i n t o ;he former class ape l i k e l y , t o be more u s e f u l f o r c o n t e n t r e p r e s e n t a t i o n t h a n t h o s e i n t h e latter.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "To A t t h e bottom of Table 3 , range and average values are given f o r those terms among t h e sets of 50 terms for which t h e term relevar~ce is defined (that is, lhose which co-occur j o i n t l y in some query-document p a i r ) .", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 23, |
| "end": 30, |
| "text": "Table 3", |
| "ref_id": "TABREF5" |
| } |
| ], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "Again the term relevance values are s u b s t a n t i a l l y d i f f e r e n t f o r t h e two classes of DV terms, and these differences are s t a t i s t i c a l l y s i g n i f i c a n t .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "Also included i n Table 3 The data already included i n Table 3 are shown i n term relevance o r d e r i n Table 4 . The output o f Table 4 ", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 18, |
| "end": 25, |
| "text": "Table 3", |
| "ref_id": "TABREF5" |
| }, |
| { |
| "start": 56, |
| "end": 63, |
| "text": "Table 3", |
| "ref_id": "TABREF5" |
| }, |
| { |
| "start": 107, |
| "end": 114, |
| "text": "Table 4", |
| "ref_id": "TABREF8" |
| }, |
| { |
| "start": 132, |
| "end": 139, |
| "text": "Table 4", |
| "ref_id": "TABREF8" |
| } |
| ], |
| "eq_spans": [], |
| "section": "T e r m v a l u e Measurements", |
| "sec_num": null |
| }, |
| { |
| "text": "0 . 00002 average high/avsrage low", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "23.93% t-test", |
| "sec_num": null |
| }, |
| { |
| "text": "874.00 t o 0,00 average high/average low 1.3 2. , -, -------. r --, -----L . -----. L -, -c -1 .", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 48, |
| "end": 94, |
| "text": ", -, -------. r --, -----L . -----. L -, -c -1", |
| "ref_id": "FIGREF14" |
| } |
| ], |
| "eq_spans": [], |
| "section": "Term range", |
| "sec_num": null |
| }, |
| { |
| "text": "-------- t ------------------- S/N", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "d --------------------c ----------------------------d -", |
| "sec_num": null |
| }, |
| { |
| "text": "a . ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "IS(TO( . D E T 4 ) ) C A T ( 'PRO'] aS(TO(.PFO)) C A T ( 'NPR \") I S ( T O ( . N P R ) > F ( F R E T U R N I SETR( .NP, 'PROP',Q) ~( T O ( . P O P N P ) ) SETR( . N P , * P R O * , Q ) t (-TO(. P O P N P ) ) S E T R ( .NS, ' D E -T * , Q ) C A T ( * A D J * )", |
| "sec_num": null |
| }, |
| { |
| "text": "C A T ( 'V", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "PP = B U L L D S ( P P ) ~( R E T u R H )", |
| "sec_num": null |
| }, |
| { |
| "text": ".", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "D I P = (AUX)(TNS P A S T ) . C A N = ( A U X ) ( T N S P ' R E S ' ) . C O U L D = (EOR!/l ' C A : : )", |
| "sec_num": null |
| }, |
| { |
| "text": ". ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "W I L L = ( A U X ) ( T t l S F U T ) .I T H E = ( D E T ) . A = ( D E T ) . A N = ( U E T ) . T H A T = ( C L I ! ! l ) )", |
| "sec_num": null |
| }, |
| { |
| "text": "V f L L A C E B U I L D S T R U C T U F E Y O U STATE P O P N P COMPLEEE!;T S T R i N C ; W A L K T O THE V I L L A G E B U l L P S T R U C T U R E Y O U S T A T E QEiP", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ")ID Y O U W A L K T O T H E V I L L A G E D L D Y O 2 W A L K T O T H E V I L L A G E S T A T E C U E S COF?PLEuZZdT STRING:' YOU U A L K TO T H E V I L L A G E B U I L D S T S U C T 3 R E DID S T A T E P F O C O M P L E Y E Y T S T R I N G a W A L K ~d T H E", |
| "sec_num": null |
| }, |
| { |
| "text": "( V P ( V", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "C O M P L E Y E N T S,Ti?I N G : B U L L D S S R U C T U R E :", |
| "sec_num": null |
| }, |
| { |
| "text": ")", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "W A L K ) ( P P ( P P ( P R E P TS)(PREPNP ( N P ( D E T THE)(N VILLAGE)))))) STATE S COMPLEME?iT S T R I N G 8 B U I L D S T R U C T U R E r (S(TYPE Q 3 E S T I I : O ) ) A A X D D D D J T E E I S E P A S T ) ( S ' J & J ( N P ( P R 0 Y O U ) ) ) ( P R E D ( V -P ( V WALK) ( P P ( P a P ( P R E P T O ) ( P R E P K P ( N P ( D E T T H E ) ( N V I L L A G E ) ) ) ) ) ) )", |
| "sec_num": null |
| }, |
| { |
| "text": "E P \\ P \\ O F P C P s ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "O T O U T P U T = P O P ( P A T H ) S S ( O T ) F ( E X A S \\ S \\ M I N )", |
| "sec_num": null |
| }, |
| { |
| "text": "S T A T E 3 . P", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "B U I L D S T S U C T U R E MvdA", |
| "sec_num": null |
| }, |
| { |
| "text": "1 E N G L I S H :", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "C O Y P L E E E N T S T R I N G : B U I L D S T R U C T U R E ( M P ( N O L T N A S . ) ( P O S P R Q F A N A H A D J N A N -? : A~J ) ( D E T C E H P L W A )", |
| "sec_num": "91" |
| }, |
| { |
| "text": "DCl3 M Y B I G A P L U R D O Y O U W A N T TO E X A M I N E T E E R E G I S T E R S ?", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "C O Y P L E E E N T S T R I N G : B U I L D S T R U C T U R E ( M P ( N O L T N A S . ) ( P O S P R Q F A N A H A D J N A N -? : A~J ) ( D E T C E H P L W A )", |
| "sec_num": "91" |
| }, |
| { |
| "text": "A S T A T E N P C O M P L E M E N T S T R I N G : D I D N O T P A R S E B U I L D S T R U C T U R E M'dA D O Y O U W A N T TO E X A M I N E T H E R E G I S T E R S ? N O 1 N P U T . S T R U C T U R E T O BE P A R S E D AS WdA A S MWA S T A T E P O S C O M P L E . M E~T I", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "S T A T E ? O S C O R P L E H S N T S T R I N S : FAMA B I J I M C E MW4. B U I L D S T F U C T U R E AS S T A T E A D J C O Y P L E M Z N T S T R I S G : B I J I M C E M W A , B U I L D S T 8 U C T U R E F A N A S T A T E D E T C O M P L E M E N T S T R I N G : C E B U I L D S T R U C T U R E B J J I H S T A T E P L T C O M P L E M E N T S T R I N G : B u r m S T~H J C T Y R Z ~W", |
| "sec_num": null |
| }, |
| { |
| "text": "A S", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "S T R I N G : M W A B U I L D S T R C~C T U R E '", |
| "sec_num": null |
| }, |
| { |
| "text": "B U I L D S T F U C T U f i E Y W A ST A'-?' E Y C O X P L E 3 , F N T S T R I N G : B U I L D . \" s T R i l C W T U F E P C AS) ( P L M X A ) )", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "S T A T E K T COMPLEMENT S T R e I N G : M W A B U I L D S T R U C T U a E S T A T E A D J C O N P L E M E N T STRING: MrdA B U X L C S T R U C T U R E S T A T E K O M C O M P L E M E f i T S T R I N G : pl !A' B U I L D S T R U C T U R E S T A T E DET G O X P L E M G N T S T R I N G : M W A B U I L D , S T R U C T U R E S T A T E P L C O M P L E Y E N T S T S I N G : M W A B U I L D S T R U C T U R E S T A T E F L T C O M P L E Y E N T S T R I N G : B U I L D S T R U C T U R E M W A S T A T E P O P U P C O M P L E X E Y T S T R I N G :", |
| "sec_num": null |
| }, |
| { |
| "text": "Example ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "E N G L I S H : , D O G P L U R D O Y O U W A N T T O E X A N Y E T H E R E G I S T E R S ? u n", |
| "sec_num": null |
| }, |
| { |
| "text": "3 F L~v C T I O N DEFINITIONS G,RAP", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "E N G L I S H : , D O G P L U R D O Y O U W A N T T O E X A N Y E T H E R E G I S T E R S ? u n", |
| "sec_num": null |
| }, |
| { |
| "text": "A X I S~T E S T~ T TYPE^, ~D C L . ) P S S E T ( @ T Y P E~, . T F ? P A~~~ DGLJ 4:KE:J I I I F '~F ' F ' O I t 1 . 1 tF'PED f V P 1 T I i I P H f T l r v T IHIIII tnBJ ~.~{P;PPD 41'UtJl 1 1PF.EP IIIITHI~{F'I,FFIO HEPl 1 1 i 1 1 1 1 1 1 ) 1 1 J NFUT STF:LJCTLIr*:C: 7 CI CL\":'*AKSTD ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "S E W [ @SI'AUX'~Q) S E T R (~S p P T h S ' I G E T F ( C T N 5 P ) ) EL7F~,Ev'TYPEC,\"C'3 F E P S E ( b P 0 ) ~s ( T Q ( , Q N P I ) F ( F R E T~R N )", |
| "sec_num": null |
| }, |
| { |
| "text": "J. I i ( s / N )~ = log F -C l o g -: -A l? E k", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "YOU W A N T T U EXAMINE THE I?'I,GISTEI?S '3 NQ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "Perceptual categories for musiclike sounds: implications for theories of speech perception", |
| "authors": [ |
| { |
| "first": "J", |
| "middle": [ |
| "E" |
| ], |
| "last": "Cutting", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [ |
| "S" |
| ], |
| "last": "Rosner", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [ |
| "F" |
| ], |
| "last": "Foard", |
| "suffix": "" |
| } |
| ], |
| "year": 1976, |
| "venue": "Quarterly Journal of Experimental Psycholoqy", |
| "volume": "28", |
| "issue": "", |
| "pages": "361--378", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Cutting, J. E., Rosner, B. S., Foard, C. F. (1976) Perceptual categories for musiclike sounds: implications for theories of speech perception. Quarterly Journal of Experimental Psycholoqy, 28 : 361-378.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Perceptual mechanisms at the first level.of speech processing", |
| "authors": [ |
| { |
| "first": ";", |
| "middle": [ |
| "A J" |
| ], |
| "last": "Fourcin", |
| "suffix": "" |
| } |
| ], |
| "year": 1971, |
| "venue": "Proceedings of the V I I t h International Congress", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Fourcin; A. J . (1972) Perceptual mechanisms at the first level.of speech processing. In: A. Rigault and R . Charbon- n e a~ , eds . Proceedings of the V I I t h International Congress o f ~h o n e t -i c Sciences, Montreal 1971..", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "Probabilistic Models for Automatic Indexing", |
| "authors": [ |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Bookstein", |
| "suffix": "" |
| }, |
| { |
| "first": "D", |
| "middle": [ |
| "R" |
| ], |
| "last": "Swanson", |
| "suffix": "" |
| } |
| ], |
| "year": 1974, |
| "venue": "ASIS", |
| "volume": "25", |
| "issue": "5", |
| "pages": "312--318", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "A . Bookstein and D.R. Swanson, Probabilistic Models for Automatic Indexing, J o w n a l of t h e ASIS, Vol. 25, No. 5, ~e~t e m b e r~~c i o b e r 1974 , p . 312-318.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Testing of a Fully Automatic Indexing-Searching System f o r Documents Consisting of Expository Text", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [ |
| "P" |
| ], |
| "last": "Dennis", |
| "suffix": "" |
| } |
| ], |
| "year": 1967, |
| "venue": "Information Retrieval: A Critical R e v i e w , G. Schecter , e d i t o r", |
| "volume": "", |
| "issue": "", |
| "pages": "67--94", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "S.P. Dennis, T h e Design and Testing of a Fully Automatic Indexing- Searching System f o r Documents Consisting of Expository Text, in Information Retrieval: A Critical R e v i e w , G. Schecter , e d i t o r , Thomps~h Book Co. , Washington, 1967, p . 67-94.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "Society for I n d u s t r i a l and Applied Mathematics", |
| "authors": [ |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Salton", |
| "suffix": "" |
| } |
| ], |
| "year": 1975, |
| "venue": "A Theory of Indexing, Regional Conference-S e r i e s in Appliea Mathematics No", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "G. Salton, A Theory of Indexing, Regional Conference-S e r i e s in Appliea Mathematics No. 18, Society for I n d u s t r i a l and Applied Mathematics., P h i l a d e l p h i a , 1975.", |
| "links": null |
| }, |
| "BIBREF7": { |
| "ref_id": "b7", |
| "title": "A Theory of Term, Importance i n Automatic Indexing", |
| "authors": [ |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Salton", |
| "suffix": "" |
| }, |
| { |
| "first": ".", |
| "middle": [ |
| "C S" |
| ], |
| "last": "Yang", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [ |
| "T" |
| ], |
| "last": "Yu", |
| "suffix": "" |
| } |
| ], |
| "year": 1975, |
| "venue": "", |
| "volume": "26", |
| "issue": "", |
| "pages": "33--44", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "G . Salton ,. C. S . Yang and C. T. Yu, A Theory of Term, Importance i n Automatic Indexing, J o u r n a l of the A S I S , Vol. 26, No. 1, January- February 1975, p . 33-44.", |
| "links": null |
| }, |
| "BIBREF8": { |
| "ref_id": "b8", |
| "title": "A Vector Space Model for Automatic Sndexing", |
| "authors": [ |
| { |
| "first": "G", |
| "middle": [], |
| "last": "E A L T O N", |
| "suffix": "" |
| }, |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Wong", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [ |
| "S" |
| ], |
| "last": "Yang", |
| "suffix": "" |
| } |
| ], |
| "year": 1975, |
| "venue": "Communications of the ACM", |
| "volume": "18", |
| "issue": "11", |
| "pages": "613--620", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "G. E a l t o n , A. Wong, and C.S. Yang, A Vector Space Model for Automatic Sndexing , Communications of the ACM , Vol . 18, No. 11, November ' 1975, p. 613-620.", |
| "links": null |
| }, |
| "BIBREF9": { |
| "ref_id": "b9", |
| "title": "Precision Weighting -An E f f e c t i v e Automatic Indexing Method", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [ |
| "T" |
| ], |
| "last": "Yu", |
| "suffix": "" |
| }, |
| { |
| "first": "G", |
| "middle": [], |
| "last": "", |
| "suffix": "" |
| } |
| ], |
| "year": null, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "C.T. Yu and G. S a l t o n , Precision Weighting -An E f f e c t i v e Automatic Indexing Method, to be published in J o w n a l of the ACM, 1376.", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "The C o r n e l l Emplementation of the S M A~T System, in T h e SMART R e t r i e v a l System", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [], |
| "last": "W I L L I A M S O R I", |
| "suffix": "" |
| }, |
| { |
| "first": "R", |
| "middle": [], |
| "last": "Williamson", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Lesk", |
| "suffix": "" |
| } |
| ], |
| "year": 1971, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "D. W i l l i a m s o r i , R. Williamson, and M. Lesk, The C o r n e l l Emplementation of the S M A~T System, in T h e SMART R e t r i e v a l System, G . S a l t o n , e d i t o r Pren-k ice -Hall, EngLewood Cliffs , NJ , 197 1 , Chapter 2 .", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "FIGREF0": { |
| "num": null, |
| "type_str": "figure", |
| "text": "", |
| "uris": null |
| }, |
| "FIGREF1": { |
| "num": null, |
| "type_str": "figure", |
| "text": "ITtportance a n d t i m e l y i n t e r e s t to t h e m e v h e r s h i p .The cast", |
| "uris": null |
| }, |
| "FIGREF4": { |
| "num": null, |
| "type_str": "figure", |
| "text": "March 1 9 7 7 , P J . k ' t j 1 F~: 1 4 t n b n n u -3 1A C L H u s i n e s s 1Cleetinq r n a , t~l q a r i a , in 1 9 7 8 ,", |
| "uris": null |
| }, |
| "FIGREF5": { |
| "num": null, |
| "type_str": "figure", |
| "text": "D o n d a l k e r , SHI F x e c u t i v e C o w m l t t e e : J e r r y Hoobs. C U k y y o m i m a t i n g C o r n n i t t e e : Stan Y e t r i c k , l t 3 Y 4 %tion t h a t t h e slate b e a c c e p t e d u n a n i~1 o u s 1~ b a s c a r r i e d , b", |
| "uris": null |
| }, |
| "FIGREF6": { |
| "num": null, |
| "type_str": "figure", |
| "text": "h e me-etinq a d j o u r n e d . Dona13 t. b a l k e r S e c r e t a r y -~r e a s u r e r , P r o -r e~~ Microfiche 5 5 : 7'", |
| "uris": null |
| }, |
| "FIGREF7": { |
| "num": null, |
| "type_str": "figure", |
| "text": "Some m e m b e r s a o n o t l f k e microfiches,3 .", |
| "uris": null |
| }, |
| "FIGREF8": { |
| "num": null, |
| "type_str": "figure", |
| "text": "", |
| "uris": null |
| }, |
| "FIGREF9": { |
| "num": null, |
| "type_str": "figure", |
| "text": "C L s e c r e t a r y -T r e a s u r e r s R e p o r t , h , U c t o b e r 1 p t i n u e d , unreal st ically l o w d u e s r a r e o f s~O , f o r the d u e s at $ 1 0 U h t i l S u c h t i m e as t h e A C L d r r i d e d * P l a t t o 0 0 j~b o d~t :", |
| "uris": null |
| }, |
| "FIGREF10": { |
| "num": null, |
| "type_str": "figure", |
| "text": "", |
| "uris": null |
| }, |
| "FIGREF12": { |
| "num": null, |
| "type_str": "figure", |
| "text": "r o p r i : l t e . Thc r c p r c s m~t a t i o n o i t h e Anowl c J y c \\ a v a i l a b l e t o s u c h a n '!rlndcrst nndcr\" s y s t g~n i s an impel-t a n t i s s u c f o r t h e s y s t c m l s d c s i g n ,~n d i s i n t i m a t e l y related t o i h c pl-ogosctl u s e s of-t h a t knowlcclge. 'I'his book i n c l u d c s n c o l l c c t l o~~ o f t h i r t e e n p;lpcrs. wri tten by some of t h e b c s t known r c s c n r c h c r s 1~110 are currcntly w o r k i n g on ~~n d e r s t a n d c r systrms. T h e p n p c r s wcrc s e l c r t c d nnorlg those presented a t a conference held in mcinory of Ja ilne C a r l~o n c l l . R t -p r c s~n t a tion and 1lndc.r~ tanding l'hc c-ontct~ts of tllc Imok arc iis f o l lnws 1 . * 1 l p o f Hcp~ c s~n t i~t i~n 1 . llimcns ions of Rcprcscnt i~t ion !kinicl G. 13obrow 2 . I\\')~:lt's i l l a, Link: P o u n d n i i o l~s for Scmnntic Mctworks I 1 1 1 1 A 'lVlmds 3 . RcTlc3ct i o n s on t h c 1:ormnl J I c s c r i p t i o n o f Rchovior J O S C F~ n. U C C~C T 4 . S y s t c m n~i c ilnc1erst:rildin~: Synthcsi s , Analysis, a n d Cont ingcnt Kno\\~rmlcclge in Spcci:11 i zccl i l n d c r s t il11tl i ng Systcms Rohcrt J. Robrow 6 J o h n S c c l y Brown Solnc Principles of Xcmory Schemata Daniel G. R01)row C;. Donald A . Yorman 6 . A Fran~c Tor Fratilcs : Notes on a Scllc~na f o r S t o r i e s Thc St r u c t u r c of Epi sodcs in Pfcmory Rogcr C . Schnnk 1 0 . Conccpts f o r R c p r c s c n~i n g Yundnnc Rcoliiy i n P l a n s Rohcrt P. Ahclson 1V. Srm:~ni i c Know1 cdgc i n i~n J c r s i :~~r i l c . r Flu1 t i p l c R c p r c s c n t a t i ons of Know1 cdgc f o r T u t o r i a l R e a s o n i n g J o h n Sccly Rrown 6 ~i c h n r d R. Burton 12. Thc Rolc o f S c m a n t i s s i n A u t o m a t i c Spcech l f n d c r s t a n d i n g Bonrric Nush-Kcbbcr 13. R e a s o n i n g From Incompl c t e X~l o r r l c d g c A l l u n C o l l i n s , Elcanor 11. ICarnock, N c l l c k c A i c l l o , I . M i l l e r As s t a t e d i n t h e b o o k ' s i n t r o d u c t i o n , t h e s e c t i o n on \"Thcory o f R c p r e s c n t a t i o n \" d e a l s w i t h gcncral issues regarding t h e rcprcscntation o f knorvlcrlge, while t h a t on lfScw afclnory ?-lodclsH e . The s c c t i o n t i t l e d \"Ilighcr L c v e l Structures\" f o c u s e s on t h e rcprcscntation o f plans, c p i s o d c s and s t o r i e s w i t h i n memory. F i n a l l y , t h e s c c t i o n on \"Semantic knowlcdgc i n l l n d c r s t a n d c r Systcms\" dcscr~ibcs o n -g o i n g work a f t h e SO1311TE, SPI~I~CIILIS and SCIIOI.AR pro j cct s n t BRN.", |
| "uris": null |
| }, |
| "FIGREF13": { |
| "num": null, |
| "type_str": "figure", |
| "text": "d c c l n r a t i v c , V S . I l c x i b l c interaction ilnlong cli f f r r r n t F a c t s , f o r procedural . Schank ' s paper r 1 O 1 i n c l udcs a cl i s'cussi 011 o n whct h c r t h c o~~g n n i z a t i o n o f human mcmory i s c p i s o d i c o r scmant i c . An cpi s o d i c mcmory o r g n~~i ;at i o n i111pl i c s t h a t h~i o~\\ r l ccljic i s s t o r c d l s s t c a p o r ;~l ly d a t e d c p i s o~l c s and cwcnts, w i t h 1 rmpornl spat i n 1 r c l a t i o n s 1 i n k i n g thcsc cvcnts. A S~m i l n I i c nlcinory organ i z i l t i o n , on t h e o t h e r ha11c1, i n v o l v c s t imc-i n v a r i n n t know1 cdgc n p c r s o n p o s s c s s c s , c . g . , \" a l l elephants a r c animals1'. A c o r . o l l a r y 01 t h e s e d c f i n i t ions j s t h a t an c p i s o d i c mcmory o r g i~n i z a t i o n f ' ,~v o~r r s tclllporal and causal c o i~n e c t i v o s (e .g. , TllliN, Rl:ASON, I:NABIJE c t c . ) , w h c r c a s a s c~n a n t j c mcmory org:lni z n t i o n u s c s r x t c l l s ivel y t h e \" I SA hicrnrchy!' ( e . g . , \"an cl e p l l a n t i s -a a n i m a l H ) . The d i scussion p r e s c n t c d i n t h c papcr on t h i s issue i s somcwhnt c o n f u s i n g s i n c e a t one p o i n t (pp. 255-256) t h e t v o t y p c s o f or.gnni z a t i o n arc c o n t r a s t e d as i f t h y -wcrc m u t u a l l y c x c l u s i~r c , w h i l c l a t e r on (p. 263) t h c p a p e r a r g u e s Tor a c o m b i n a t i o n o f the n o t i o n s of s c m a n t i c and e p i s o d i c mcmory. in cithcr c a s c , Schnnkls work c e r t a i n l y makes a c o n v i n c i n g argumcnt i n rnvor o f a n c p i s o d i c monory o r g a n i z a t i o n by sllowing how jt c a n h c usccl t o r c y r r s c n t thc mcnning of a p a r a g r a p h .", |
| "uris": null |
| }, |
| "FIGREF14": { |
| "num": null, |
| "type_str": "figure", |
| "text": "Crl --,prc and ----E x t c~l s i o n s o r R c j~r e s c n t n t i o n o f l i~~o \\~l c c l g c P a r :~J i g~n s I , , --, ----, , , ---------Scvcral p a p c r s , includjng some t h a t were mcntioncd i n t h c p r c v i o u s scction, criticize, r c f i n c , o r cxtcncl o n e of t h e csisting p a r a d i g m s Cor t h e rcprcscnt:it i on o f knorul ctlgc. l'hc most n o t : l l~l e c s a~s p l c alaong tllosc i n i l l i s calcl:ory i s I\\'oodst p a p r r r 2 3 i c r i t i c i z e s nlany ( m i s ) u s c s o i s c m i~n t i c n c t w o r k s by p o i n t i n g o u t s i t u a t i o n s w h c r c t h c i r sc'mantics a r c p o o r l y dcf i n c d a o r inconsistent. P a r t i c u l a r a t t c n t j o n i s p a i d t o t l~c rcprcscntation o f qunntlficntion and t h a t o f rcla1 i v e c l n~i s c s . A s many o f tI1c r c n d c r s i i n d o u h t c d l y kllow, )Iinsky1s jnflucntinl p a p e r i n t r n d u c Sng I'framcs\" 11 5'1 p r o v i d c s m o r e of an i d c o l o g ) * t h a n a t h e o r y Tor rcprcscntjng knoxlcclge. Kuigcrs in [ G I :irgucs in i i l v o r of a n~~n b c r or p r o p c r t i c s Eramcs s h o u l d h;luc, such ;IS t h e", |
| "uris": null |
| }, |
| "FIGREF15": { |
| "num": null, |
| "type_str": "figure", |
| "text": "o J r s c r i b c all o b j e c t o r s i t u a t i o n t o varying d r g r c e s of d e t a i l , t h e ability t o h c i n s t a n t i ; l t c d a n d t h e :~I~j l i t y t o 11antlIc s a~n l l p c r l u r b n t i o n s o f cxpcrted i n p u t d a t a ui t h o u t m a j o r r a i l u r e s . lie illustrotcs thq dcsirabjlity of t l l c s e i c a t u r e s w i t h a s i m p l e example of o b j c c t r c c o g n i t j on. The s e c o n d h a l f of iVinogrndls paper m :~k c s an a t t c m p t t o s y n t h e s i z e d c c l a r a t i v c a n d proccdurnl aspects of a rcprcsentntion. Ilis proposal is hnscd on fr:uacs :~n d u s c s a g c n c r a 1 j z ; l t i o n (1%) r : i t i o n s on lhosc ol~jc~cts. U:my of t h c iclcns i n TSJ and 1 7 1 h n v c b c c n incorporated i n KRL r161, a s developed by D. Bobrow and IVinogrnd. I T 1 . Rcprrscnt i n p Dificrcnt K i n d s o f Know1 cdgc , -------------------, , , -, , , , --Tn Format ion cnt c r i ng an unclcrr.~ n n d c r sys t c n~ may h a v c many d i f f c r r n t V o r m s V , i. o . it may be codcd as p h o t o g r a p h s o r 1 i n c c l r n w i n g s , sinplc s e n t rnces o r p a r a g r a p h s o r c v c n coirpl ct-c s t o r i c s . Florcovcr , i t may h n~~c d i f ' f c r c n t \"corlt cnt'! i . c . i n v o l v c a f a i r y i n l c w o r l d o f 1;inl;s a n d t l r n g o n s , ab l o c k s v:orId of' c~i b c ' s a n d pyl-amids, a s o c i a l , m c n t a l o r p l l y s i c a l worltl. Onc j~r p o r t n n t a s p e c t o i t h e r c p r c s c n t n t j o n p r o b l c~n is t h c d c r i n i t i o n o f a collection o i k n a~c l c d g c , dcfincd by n r e s t r i c t i o n on i t s rorm a n d / o r c o n t e n t , and t h e i n v c s t i g a t i o n of t h c nclcquncy o f a p a r t i c u l a r r c p r c s c~~t o t i o n .", |
| "uris": null |
| }, |
| "FIGREF16": { |
| "num": null, |
| "type_str": "figure", |
| "text": "s ~n c n t i onbd c r l i c r , lroods ' pilpcr d o c s (1 i s c u s s t h c r e p r cs c n t a t i o n o r q u a n t i f i c a t i o n i n t c r m s o r s c m ; i n t i c networks, w h e r e t h c form o f tllc l \\ n o v l c d g c i n v o l v e d i s prcsumnl?ly ( f i r s t o r d e r ) P r c d i c a t c Ca7 cul u s a n d t h e c o n t c~l t i s ~i~l c o n s i -r ; i i ncd . I t a1 s o d i s c u s s e s t h c rcprcscntat i o n of' r c l a t i v c clauses and conlplcx s d n t c n c c s 1111cre t h e form i s f l n t u r n l l a n g u a g e a n d t h e c o n t e n t i s , a g a i n , u n c o n s t r a j n c d . R u m e l h a r t I s p a p e r 83 j s p r i l n a r i l y c o n c c r n c d w i t h tllc d i s c o v c r y o f s t r u c t u r c u n d e r l y i n g s i m p l e s t o r j c s . The s t r u c t u r e j s d c r i n c d i n t c r m s o f n p1lr;lsc s t , -u c t u r c gr;lalm;lr w i~h s c l~n ; u l l i c ~u l cs associated t o e a c h product. i o n . Thc p n p c r r c r t a i n l y f o l l o w s t h c g c~l c r s l trcncl t o \\~; i s c l s s t l i d y i n g 1 i n g u i s t i c u11 i t s 1:lrgcr t11n11 sent c n c c s , s u c h a s p n r n g r a p h s , d i n l o g u c s o r st o r i c s . Whcthcr t h c methodology u s c d ( i n p a r t i c u l ;lr, p l l r a s c structure gr:ln1111:1rs) w i l l b c f o u n d n d e q u a t c f o r t h c d c s c r j l t i o n of s t i -~i c t u~c i n s t o r i e s rcmnins t o b e s c c n . Sch:lnk 1 9 1 d c a l s m a i n l y w i t h t h e p r~b l c m of c o n s t r u c t i n g a s t r u c t u r e o f c a u s a l l y -1 i~i k c d a c t i o r l s nncl ( -l~:~n j ; c s o r s t n l c s ~cprcsentation and Understanding 23 ( c p i s o d r s ) from n p n~* a g r i i p h . lShen c p i s o d c s a r c used t o m a k e scnsc o f ncw i n p u t s i n o f t e n -c x p c r i c n c e d s i t u a t i o n s , t h c y a r c c a l l c t l \" s c r i p t s \" . The pnpcr cnds w i t h a b r i e f inti-oduction of scripts. Florc d c t : l i l s :{bout them can be f o u n d i n more r c c c n t pub1 i cations by Sch:jnk and hi,s students, c . g. r l 7 , 1 8 3 . R~~~n c I l~n r t~s and Schm1k1s work a r c r c l n t c d i n t h a t t h e y b o t h a i t c~~p t t o d c l inc t h o s t r u c t u r e o f a collection o f knowledge 1 i m j t cd w i t h r c s p c c t t o form ( s t o r i e s Tor Rulnclhsrt ,p a r a g r a p h s Tor Schank) and u~~c o n s t r a i n c d w i t h rcspcct to c o~l t o l t . \\ f o r c o v e r , b o t h papcrs a g r e e t h a t t h e d n d c r l y i n g rcprcscntntion u s e d must involve c a u s a l l y -l i n k e d c v c n t s , and t h c c a u s a l c o i~n c c t i v c s t h c y employ a r c similar. A b e l s o n t s p a p c r i s conccrncil w i t h the r c p r c s c n t a t i o n of \" i i i u~~~l n n c r c n l i t y f l i n v o l v i n g s o c i a l a c t i o n s . l'hc a p p r o a c h h e f o l l o~~s i s t o p o s t u l a t e a number o f p r i m i t i v e s t a t e s and a c t i o n s for a c h i e v i n g t h e s e s t a t e s , jn terms of which h o p e f u l l y a l l s i m p l e s o c i a l b c h i~v i o w can be d c s c r i b e d , The discussion of .the , p r i m i t j v c s is q u i t c t h o r o u g h , b u t t h c cxainpIcs g i v c n do not provjde.sufficient cvidcncc that t h c p r i m j t i v c s p r o p o s c d a r c i n f a c t clcscript i v c l y a d c q u a t e . ~lbC?l s o n ' s work i s complcmcntnry to S c h a n k l s i n s c v c r e l rcspccts and t h e r e i s more rcccnt j o i n t work on t h c s u b j e c t C193. I V 1 . -O n s o j ng P r o -a-w--j-ccts i n v o . l v i~ -IJndcrstnncler asterns --, Thc last t h r c c papcrs of t h c book discuss p a r t j c u l a r p r o j c c r s i n v o l v i~~g t h e d e s j g n and jrnplcnicnt*atjon o f u n d c r s t a n d~~ s y s t e m s . ~cprovcntation and Understanding 24 I11 1 d c s c r i b c s t h c s c o p c , h a s i c ~l l c~t l~o d o l o g y , :ind :ic]~jcvcmcnt s o f S I I , n k n o~~r l r d g c -l~:~s c~c l c o m p u t r~-;~i d c d i n s t r u s t i o n ( C A I ) by a s k i n g q u c s t i o n s , :~nswcr.ing q c~c s t i o n s a n d lctting 11im t r y o u t Ilis iclcns. Of p a r t i c u l a r i n t c r c s t t o c o a p u t n t i o n n l 1 inguists ! I sLo111d bc t 11c s e c t i o n dcscr i b i n g t l~c s c~n i n n t i c g r n n i l~~a r \" dcvc1opcd by nurton t o 11:lndIc t h c typcs o f s c n~c n c c s c?xpcctcd d u r i n g a d i a l o g~i c on c l c c t r o n i c c i r c r~i t s . Nash-Kcl)l)cr I 1 2 1 p r o v i d c s a n o~~c r v i c w o f illc HBN SI)~~l~ClIJ,IS p r o j c c t i n t h c c o n t r x t o f a d i s c~i s s i o n on t h e u s c o C s c m n~~t i c knowlcdgc f o r s p c c c h umclcrst a n d i n g . F I 1 , [ 137 discusses some o f t h c i n r c r c n c c r u l c s jlnpl cacntocl o r b p i ng eonsidcrcd !'or j~~~p l c l n c n t n t i o n by the SCIIOJ,AR p r o j c c t whose a i m i s t o d c v c l o p a knowJ cdgc-based C A I systcln 1 h a t t c ; l c h r s geography . Tllc rcadcr. m a y f i l i d many o f tlrc r u l e s s t a t e d i n t h e pnpcr c o m l~l c t e l y r c a s o na b l e a n d y e t q u i t e s h a k y from a logical p o i n t o f v i e w . For example, one r u l c (the u n i q n c n c s s asyulipt i o n ) s t a t c s t h a t i f only one t h i n g is f o~i n d , i t c a n bc nsslimcd t h a t i t constitutes a complcte s e t . Thus i f sonlco~lc k~~o w s o f o n l y onc c i t y c n l l c d f l S 1~r i n g r i c l d \" a n d l o c o t c d j n ? l :~s s a c l~~~: ; s c t t s , h c c a n u s r I he uniqueness nssumpt i on t o r c p l y \"no\" t o \"1s Springfj c l d i n Kc~ltucky?\" e v e n t h o u g h t h c r c 11iny w e l l be such a c i t y . 'l'he p a p e r s i n t h i s s c c t j o n c o n s t i t n t c an i m p o r t a n t complcmcnt t o t h e r c s t o f tllc book w l l i c l~ o f t e n i r l v o l v c s d j s c u s s i o~i s t l l a t arc t o o far rcmoVcd from t h c r c n l i t y o f I i l l l p l r~n~~l t c d ( o r imp1cmcnt;iBJ c: f r c~r c s c~~t a t i o n and u n J c r s t a n d c r s y s t c m design. For more i n t r o d u c t o r y discussions, t h e rcBdcr is r c f c r r c d to / I 4 1 o r Schank a n d Colby L 2 ICinograd, T. \"Five L e c t u r e s on A r t i f i c i a l Tnt eJ ligcncc\" S t a n f o r d AI-h4cmo 2 4 6 , Scptcmbcr 1974. 15. Minslcy, ? I . \"A Framework f o r Rcprcscnt i n g Know1 edge\" i n Winston P . ( E d . ) The ---p a -Psycholog) ---..--o f Computcr Vjsion, -.-----McGrau Hill, 1975, 16. Robraw, TI. and Winograd, T. \"A KRL U s c r l s Manual\" (uilpubi j s l~c d )", |
| "uris": null |
| }, |
| "FIGREF17": { |
| "num": null, |
| "type_str": "figure", |
| "text": "A 1 i;youp \"SAM ---a Story Unclerstnndcrff Yalc University, Dcpt. o f Computer Science, A u g u s t 1975. 1 9 . S c h a n k , R. and Abelson, R . \" S c r i p t s , PI a n s and Know1 cdge\" P.roccodings IJCAI, p p . 151-157, Scptc~nbcr 1975. 20. Schank, R. and Colby, K. ( E d s . ) Computer Vodels .-of Thougllt -and Language, Frccman, 1973.", |
| "uris": null |
| }, |
| "FIGREF18": { |
| "num": null, |
| "type_str": "figure", |
| "text": "at t h e Urban Life Center, Columbia, Maryland. T h e conference was one i n a s e r i e s c a l l e d \"Communicating by Language\", sponsored by t h e National. ~n s t i i u t e of Child Health and Human Development ( N I C H D ) . These are 1 9 papers, divided into 3 major sections, viz. I The development of speech i n man and child I1 Language without speech (dealing with s i g n language) I11 Phonology and language Some papers are followed by comments of one of the p a r t i c i p a n t s each p a p e r o r c o h e r e n t group of papers i s followed by a summary of fhe open discussion. A separate IVth section of the book contains r e f l e c t i o n s on the conference by Ira J . Hirsh. Refe-The R o l e of Speech in Language rences a r e presented a t t h e end of each p a p e r . The e d i t o r s have provided a name index and a s u b j e c t index a t the end of t h e book. Many l i n g u i s t s and p s y c h o l i n g u i s t s take i t f o r granted t h a t language can be s t u d i e d without studying speech. Likewise many speech r e s e a r c h e r s seem t o work from rhe view t h a t the p~o d u c t i o n and p e r c e p t i o n of speech can be s t u d i e d without s~u d y i n g language. This s i t u a t i o n l e a d s Alvin Liberman t o I I s t a t e i n h i s \" I n t r o d u c t i o n t o the conference\" t h a t our t o p f c --t h e r o l e of speech i n language--is not an e s t a b l i s h e d one; no one has made i t t h e d i r e c t and primary o b j e c t of h i s r e s e a r c h . 1 1 Although t h i s statement i s perhaps too c a t e g o r i c a l , i t c e r t a i n l y i s v a l i d f o r most of t h e f i e l d . (An obvious exception, t o my mind, i s among o t h e r s Professor Lindblom of the University of Stockholm, who s y s t e m a t i c a l l y explores t h e explanatory value of q u a n t i t a t i v e models of speech production and perception i n phonology, e . g . Lindblom 1 9 7 2 , 1 9 7 5 ) . The o r g a n i z e r s of the conference, Kavanagh and Liberman, have taken c a r e t o s e l e c t well-known r e s e a r c h e r s with d i f f e r e n t backgrounds and d i f f e r e n t i n t e r e s t s t o d i s c u s s t h e v a r i o u s problems which may be derived from t h e c e n t r a l q u e s t i o n : \"do we i n c r e a s e our understanding of language when w e t a k e i n t o account t h a t i t i s spoken?\" T h e r e s u l t i n g t e x t s make i n t e r e s t i n g r e a d i n g , although one w i l l look i n v a i n f o r a convincing answer t o the i n i t i a l q u e s t i o n . D i f f e r e n t i n v e s t i g a t o r s have d i f f e r e n t opinions and the p r e s e n t s t a t e of knowledge does n o t seem t o make i t The R o J~I of Speech in Language p o s s i b l e t o settle the m a t t e r . In most papers specialist knowledge i s freely intermixed w i t h s p e c u l a t i o n , and i t i s not always e a s y t o t e l l the one f r o m the o t h e r . The discussions g e n e r a l l y serve more to con-tinrle speculation than t o criticize i n d e t a i l each other's t h i n k i n g . These remarks a r e not meant a s a criticism of the conference and i t s proceedings. They i n t e n d t o g i v e an i n d i c a t i o n , however, of the s t y l e of this book, and a warhing t h a t one w i l l n o t find here a thorough d i s c u s s i o n of empirical d a t a o r e x p l i c i t , t e s t a b l e theories, t h a t could be of use in more p r a c t i c a l l y o r i e n t e d work. I n s t e a d one f i n d s a number of i n s p i r i n g e x p o s i t i o n s of such d i v e r s e topics a s similarities and dissimilarities between human and a n i m a l communication systems, t h e evolutionary connections between language, speech, and tool-making, t h e primacy of production o r perception i n the phylogenesis and the ontogenesis of speech, the primacy of signs or speech i n t h e e v o l u t i o n of language, t h e a r t i c u l a t e s t r u c t u r e of signs in those who have s i g n language as t h e i r first language, t h e origins of phonological change, and the p a r a l l e l s i n phonological and other l i ng u i s t i c o r g a n i z a t i o n of language. Below I w i l l make a few remarks on a few selected topics: a) The evolutiorl of speech and language b ) Spoken language and sign language c) Innate f e a t u r e d e t e c t o r s d ) The absence of prosody I w i l l not attempt t o cover i n t h i s review all p a p e r s i n t h e book.", |
| "uris": null |
| }, |
| "FIGREF19": { |
| "num": null, |
| "type_str": "figure", |
| "text": "Ro-E-' of S p e e c h in Lanyuayc transformational grammar which f o r m a l l y incorporates a memory is n e c e s s a r y . A s far a s I u n d e r s t a n d h i s r e a s o n i n g t h i s i s s o because in making a p a r t i c u l a r c h i p one h a s t o keep two t h i n g s in mind, b o t h t h e l a s t c h i p t h a t has been made and t h e f i n a l form of t h e t o o l . It seems t o me, b w e v e r , that i n order t o ~i v e h i s argument i t s f o r c e i t s t i l l has t o be shown t h a t 3 t h e r e i s a fundamental d i f f e r e n c e i n t h e n e c e s s a r y complexity of u n d e r l y i n g m e n t a l s t r u c t u r e s between L e v a l l o i s toolmaking and many forms of g o a l -o r i e n t e d b e h a v i o r we f i n d in higher a n i m a l s . Liberman a l s o s u g g e s t s that t h e final c r u c i a l s t a g e i n t h e e v o l u t i o n of human language would a p p e a r t o be t h e development of the b e n t two-tube s u p r a l a r y n g e~l v o c a l t r a c t o f modern man, w h i c h a l l o w s i t s p o s s e s s o r s t o generate a c o u s t i c s i g n a l s ehat (1) have very d i s t i n c t a c o u s t i c p r o p e r t i e s and ( 2 ) a r e e a s y t o produce, being a c o u s t i c a l l y s t a b l e . R e c o n s t r u c t i o n s from f o s s i l s t e l l him t h a t t h e Neanderthal hominids had t o do without t h i s a s s e t , and t h e r e f o r e p r o b a b l y r e t a i n e d a cormunication s y s t e m w i t h a mixed p h o n e t i c l e v e l that r e l i e d on both g e s t u r a l and k oc a l components. A t t h i s p o i n t t h e reader p a r t ic u l a r l y f e e l s t h e need f o r a n e x p e r t c r i t i c i s m of t h e v a l i d i t y of %uch r e c o n s t r u c t i o n s .Bn SPOKEN LANGUAGE AND S I G N LANGUAGET h e question whether speech o r g e s t u r a l c o m u n i c a t i o n has been more i m p o r t a n t i n the e v o l u t i o n of human language came up several times during t h e c o n f e r e n c e . I n r e a c t i o n to M a t t i n g l y ' s The \u00ff ole of Speech i n Lanryudye i d e a t h a t \"speech exemplifies a thoroughly and p e c u l i a r l y human kind of knowing\" Hewes commented that the depigmentation of the v o l a r s k i n would i n d i c a t e the antiquity of nonvocal cormn~nication. I n d i r e c t support f o r t h i s supposed a n t i q u i t y of g e s t u r a l communication comes from some f a s c i n a t i n g s t u d i e s of American S i g n Lansuage (ASL), a c c o r d i n g t o B e l l u g i and Klima a f u l l -f l e d g e d language of i t s own, and n o t a d e r i v a t i v e o r degenerate form of w r i t t e n o r spoken E n g l i s h . Stokoe argues f o r t h e a n t i q u i t y o f sign language from a p o s s i b l e p a r a l l e l between ontogeny and phylogeny. I t appeatrs t o be t h e .case t h a t t h e i n f a n t w i t h deaf p a r e n t s , l e a r n i n g ASL a s its first language, begins p u t t i n g wordlike signs i n t o s e n t e n c e l i k e s t r u k t u r e s a t an e a r l i e r age than t h e c h i l d making two-word o r three-word sentences i n speech.", |
| "uris": null |
| }, |
| "FIGREF20": { |
| "num": null, |
| "type_str": "figure", |
| "text": "and Klima have s t u d i e d sign language from histor i c a l changes i n t h e form of s i g n s , i n s h o r t term memory experiments, by analyzing a c o l l e c t i o n of \" s l i p s of t h e hand1', sad by comparing A m e r i c a n Sign Language w i t h Chinese S i g n s , i n a l l c a s e s w i t h profoundly deaf p e a p l e who use s i g n language as their primary f o r m of communication. They show t h a t s i g n si n ASL a r e n o t s i m p l y s i g n a l s which d i f f e r uniquely and h l i st i c a l l y from one another b u t a r e , rather, h i g h l y coded units.They a l s o provide evidence that grammatical processes bear t h e marks of t h e p a r t i c u l a r transmission system i n w h i c h t h e lan-guage developed. This seems t o b e donfirmed i n ~u t t e n l o c h e r ' s Thc R o l c of Speech in Lany uaqe c o n t r i b u t i o n , comparing t h e encoding of s p a t i a l r e l a t i o n s i n ASL and n a t u r a l language (= spoken American English) I t i s too e a r l y t o draw any d e f i n i t e conclusions from t h e s e s t u d i e s of sign language on t h e interdependence of n a t u r a l language and speech, as the s t r u c t u r e of s i g n language i s only beginning t o be understood. But i t i s c e r t a i n l y o f much i n t e r e s t t o students of language behavior t h a t t h e human p e r c e p t u a l and c o g n i t i v e systems appear t o b e so f l e x i b l e t h a t profoundly deaf people may develop v i s u a l communication systems among themselves which, i f n o t equal i n e x p r e s s i v e power and speed of communication t o n a t u r a l spoken languages, a t l e a s t come c l o s e t o them. Further comparisons between t h e syntax of n a t u r a l spoken languages and s i g n languages may l e a d t o more c a u t i o n in i n t e r p r e t h g c u r r e n t i d e a s about what i s and what is n o t i n n a t e i n our l i n g u i s t i c a b i l i t i e s . S i m i l a r l y comparisons between the e f f i c i e n c y of speech p e r c e p t i o n and t h e e f f iciency of v i s u a l s i g n p e r c e p t i o n might w e l l make us wonder whether speech p e r c e p t i o n i s as s p e c i a l a s some t h e o r i s t s l i k e t o make us b e l i e v e . C I INNATE FEATURE DETECTORS The i d e a t h a t speech p e r c e p t i o n i s mediated b y , p o s s i b l y i n n a t e , speech s p e c i f i c f e a t u r e d e t e c t o r s w a s given c o n s i d e r a b l e a t r e pt i o n i n t h e conference. This idea supported M a r l e r ' s extrapol a t i o n from i n n a t e a u d i t o r y templates i n b i r d s to. i n n a t e a u d i t o r y templates i n humans. Studdert-Kennedy provides a The Role of Speech in Lilnyuayc careful s u r v e y of t h e current e m p i r i c a l evidence concerning the p e r c e p t u a l processing of consonants and vowels, from which he concludes that t h e \"human c o r t e x i s supplied with sets of a c o u s t i c d e t e c t o r s tuned t o speech, each i n h i b i t e d from output t o t h e phonetic system i n t h e absence of c o l l a t e r a l r e s p o n s e i n o t h e r detectors\". Cutting and Eimas p r e s e n t evidence t h a t such f e a t u r e d e t e c t o r s a r e i n n a t e . Eimas has shown that v e r y young i n f a n t s , one month bnd four months of a g e , can d i s c r i m i n a t e much better between d i f f e r e n t speech sounds t h a t belong t o d i f f e r e n t phonemic c a t e g o r i e s than between d i f f e r e n t speech sounds belonging t o the same phonemic category i n a d u l t speech. One m a 7 concur, however, w i t h t h e doubt expressed by Hirsh i n h i s r e f l e c t i o n s on the conference whether E i m a s ' s data a r e about speech o r about g e n e r a l a u d i t o r y p e r c e p t i o n . One may f e e l s i m i l a r doubts about t h e i n t e r p r e t a t i o n Eimas and Cutting g i v e t o t h e d a t a s t e m m i n g f r o m t h e s e l e c t i v e a d a p t a t i o n p a r a d i g m , introduced i n speech perception s t u d i e s by Eimas and Corbit i n 1 9 7 3 and s i n c e then used by a n i n c r e a s i n g number of i n v e s t i g a t o r s . I n select i v e a d a p t a t i o n s t u d i e s i t i s shown t h a t repeated s t i m u l a t i o n w i t h a p a r t i c u l a r a c o u s t i c c o n f i g u r a t i o n , f o r i n s t a n c e a s y ll a b l eb a , may change t h e response d i s t r i b u t i o n i n a phoneme i d e n t i f i c a t i o n task, f o r i n s t a n c e t h e binary f o r c e d choice between baand measured w i t h s t i m u l i . taken f r o m t h e a c o u s t i c continuum betweenba and . In this case the number o f Eresponses would i n c r e a s e a t t h e c o s t of t h eba-responses. T h e", |
| "uris": null |
| }, |
| "FIGREF21": { |
| "num": null, |
| "type_str": "figure", |
| "text": "L G E B M I C PARSING OF CONTEXT-FREE LANGUAGES I. Introduction For many y e a r s syntactic analysis and the theor;of formal languages have d e v e l o p e d in a parallel, but not closely r e l -t e d , fashion. The work described here is an effort t.0 r e l a t e these areas by applying the tools of formal power series to the p-iroblem OF parsing. This paper presents an algebraic technique for parsing a b r o a d class of context-free grammars. By parsing we mean the process of determining whether a string of terminal s y m b o l s , 1, is a m e m b e r of the language generated b y grarnmar G i . . , is x e L ( G ) ? ) and, if it is, finding all derivations of x from the starting symbol of G. We hope that posing t h e parsing problem i n purely algebraic terms will provide a basis for examination and comparison of parsfng algorithms and grammar classes. Section 11 presents a n overview of t h e a l g e b r a i c p a r s i n g process. It provides a general notion of how the method works w i t h o u t going into detail. Section 111 contains the algebraic preliminaries and n o t a t i o n a l eonventions needed in order to describe the parsing method precisely. The formal presentation of the parsing method and the proof of correctness form Section IVI Section V contains some interesting special cases of the theorem and presents some examples 0-f p a r s e s .", |
| "uris": null |
| }, |
| "FIGREF22": { |
| "num": null, |
| "type_str": "figure", |
| "text": "11. -Overview of t h e a l g e b r a i c p a r s i n g , p recessThea l g e b r a i c p a r s i n g formalism d e s c r i b e d here i s a p p l i c a b l e to a l l c o n t e x t -f r e e grammars G = <vN, vT9 P , S > e x c e p t t h o s e t h a t c o n t a i n p r o d u c t i~n s ~f the form A B where A and B are b o t h n o n t e r m i n a l s , o r e r a s i n g rules s u c h as A -p e . The p a r s i n g process c o n s i s t s first of constructing (on the b a s i s of t h e grammar G); a p o l y n o m i a l and a f u n c t i o n d e f i n e d on p o l y n o m i a l s . A p a r s e of x i s o b t a i n e d by r e p e a t e d a p p l i c a t i o n s of the f u n c t i o n t o a p o l y n o m i a l P(x). The p r o c e s s has t w o features w o r t h y of n o t e . First, i t p r o d u c e s a l l parses of x i n p a r a l l e l . Second, t h e p r o c e s s of c o h v e r t i n g a grammar into the r e q u i r e d a l g e b r a i c form is s t r a i g h tf o r w a r d and d o e s n o t alter the s t r u c t u r e of t h e grammar. This p r o p e r t y , t h e p r e s e r v a t i o n of grammatical s t r u c t u r e , i s p a r t i c u l a r l y i m p o r t a n t i n areas such a s n a t u r a l language a n a l y s i s where t h e s t r u c t u r e t h a t a grammar p r o v i d e s i s as i m p o r t a n t as the l a n g u a g e i t g e n e r a t e s . The p o l y n o m i a l s w e w i l l u s e have t e r m s of t h e form ( Z , A ) , where Z is a string a v e r a a extended a l p h a b e t and A represents a s e q u e n c e of p r o d u c t i o n s of G. T h e process b e g i n s w i t h a p o l y n o m i a l of ordered p a i r s r e p r e s e n t i n g X , t h e s t r i n g t o b e p a r s e d . A f u n c t i o n i s r e p e a t e d l y applied t o t h e poJvnomia1; t h e number of a p p l i c a t i o n s nacessary i s bounded by. the i n p u t l e n g t h . I f the r e s u l t i n g p o l y n o m i a l c o n t a i n s a t e r m ( S , A ) w h e r e S i s t h e s t a r t i n g symbol i n G , t h e n A r e p L r e s e n t s t h e p r o d u c t i o n sequence u s e d i n g e n e r a t i n g x E s o m S . I f no such p a i r o c c u r s , then x i s not i n L ( G ) , and i f multiple p a i r s I occur (S h l ) , ( 5 ' A 2 ). . . then x is ambiguous and the A s s p e c i f y t h e several p a r s e s .", |
| "uris": null |
| }, |
| "FIGREF23": { |
| "num": null, |
| "type_str": "figure", |
| "text": "is a set ( t h e carrier) znd ' is an associative binary operation. Similarly, a monoid i s a t r i p l e c o n s i s t i n g of a set, an operation and a t w o -s i d e d identity (e.g., s , ) We will feel free to denote a monoid or semigroup by its c e r r i e r . * F o r any set V , V denotes the f r e e monoid generated by V; * * + V = <V ,concatenation,n>. Similarly, V d e n o t e s the --f r e e semigroup + generated by. r; V+ = <V , concatenat ion). We denote the length of a * + string X in 7 or V by 1x1. For an arbitrary alphabet V, we define = E ;~V~V I . The f r e e half-group g e n e r a t e d by V , H ( V ) , is d e f i n e d t o be the monoid generated by V u 9 together'with the relation aa = 1, where 1 is the monoid identity and a s any element of V. N o t e that i n H(V) the elements of 7 are left inverses but not right inverses of t h e co.rresponding e l e m e n t s of V. W e If T = <~, * , 1 > and Q = <~, + , 0 > are m o n o i d s , we deno.te by T Q t h e p r o d u c t monoid <T y Q,@, (1;0)>. The carrier of T Q is the cartesian p r o d u c t T Q and t h e operation @ i s d e f i n e d t o b e the component-wise operation of T and 0: A s e m i r i n g i s an alzebraic s y s t e m < S , + , ,O> such t h a t < S , + , O > is a commutative monoid, <S,m> is a semigroup, and t h e o p e r a t i o n d i s t r i b u t e s o v e r +: am.(b+c) = a * b + a e c , (a+b)*c = a -c + b * c . A semiring i s commutative i f t h e o p e r a t i o n i s commutative, A semiring with i d e n t i t y i s a s y s t e m <~,+;,0,1> w h e r e < s , + ; , O ) is a monoid. The semirings used in this paper are commutati~re and have i d e n t i t i e s . Furthermore, i n each case the additive i d e n t i t y is a m u l t i p l i c a t i v e zero: 0 -x = x -0 = 0 . The b o o l e a n sem%ring B c o n s i s t s of t h e c a r r i e r {0,1] under the c o m r n~t a t~v e o p e r a t i o n s + and * , where 1 -1 = l + x = I. and 0+0 = O * x = 0 for a l l x E I0,l). For an arbitrary monoid M we d e n o t e by R(M) the b a n i r i n g of polynomials described as follows : 1) Each t e r n i s of the form ca where c E B (the boolean s e r n i r i n g of coefficients) and rx E M. 2 ) Each polynomial i s a formula sum ( u n d e r +) of a finite number of terms. 3) A d d i t i o n and m u l t i p l i c a t i o n of terms is d e f i n e d as f o l l o w s : a) b u + crx = (b -f-c) a b) (ba) (cB) = (be) b P ) . Addition a,nd multiplication of p o l y n o m i a l s i s p e r f o r m e d i n the usual manner c o n s i s t e n t with 3).", |
| "uris": null |
| }, |
| "FIGREF24": { |
| "num": null, |
| "type_str": "figure", |
| "text": "v a t i o n ' , meaning t h e height of the c o r r e s p o n d i n g d e r i v a t i o n t r e e o r t h e l e n g t h of the longest path from the r o o t t o the frontier of t h e t r e e . T h e h e i g h t of a d e r i v a t i o n C w i l l be d e n o t e d by h ( C ) . ' S i n c e d e r i v a t i o n ' w i l l always mean ' l e f tmos t d e r i v a t i o n 1 i n the s e q u e l , the f o l l o w i n g assertions h o l d : Assertion 1: A d e r i v a t i o n i s of h e i g h t 0 if and only if it is of length 0 . A d e r i v a t i o n i s of h e i g h t 1 if and o n l y if i t is of length 1. A s s e r t i o n 21 L e t G b e a proper c o n t e x t -f r e e grammar, and G A -9M where IGliO. Then A i s of h e i g h t l e s s t h a n o r e q u a l t o ] M I .", |
| "uris": null |
| }, |
| "FIGREF25": { |
| "num": null, |
| "type_str": "figure", |
| "text": "L e t G = <VN, VT-, P , S> b e a c o n t e x t -f r e e grammar, I an th i n d e x s e t f o r P , and l e t the j p r~d u c t i o n of G b e L e t -jr b e a d e r i v a t i o n jr A -P i of h e i g h t n + 1. Then and and for a l l i, 1 i \" m , is a derivation of height n or less.", |
| "uris": null |
| }, |
| "FIGREF26": { |
| "num": null, |
| "type_str": "figure", |
| "text": ". . , , as the index set I. Each t e r m of a polynomial from R ( H * I*) consists of an element from H I* t c g e t h e r w i t h a coefficient from t h e b o o l e a n semiring B. The elements of H -I* will be t h e basis f o r calculating the parses of a string A . The elements of H will interact t o determine i f a product of t e r m s c h a r a c t e r i z e s a derivation. If so, the associated element og I * 3 s t h e sequence of production indices or\" the derivation. The following notational conventions will b e observed. i, j, k m, n E -N, (set of natural numbers)* I S , g , , v will denote functions. F a r t h e f u n c t i o n g, IV. An a l g e b r a i c p a r s i n g theorem Theorem ( v e r s i o n -1): Let G = <VN, vT , S, P) be a proper contextfree grammar. Then there exist homomorphisms L,, g , and (5, 2\" * and a s p e c i a l polynomial p E R r I ) such that for every T X cz VT' X = XI --\u2022 X , . Xi ' VT, contains a term A if and o n l y if A is a leftmost d e r i v a t i m of x f r o m S .", |
| "uris": null |
| }, |
| "FIGREF27": { |
| "num": null, |
| "type_str": "figure", |
| "text": "g b e a n a r b i t r a r y exhaustive division of V: T h e c o n s t r u c t i o n i s most economfcal w h e n V and V are d i s j o i n t , but 1 2 t h $ s is n o t r e q u i r e d . T h e f u n c t i o n v i s t h e homomorphism induced by the f o l l o w i n g : * v(a) = ( a , A ) , a E V and fl i s t h e i d e n t i t y i n I . Since v is a homomorphism, v ( A ) = A .The function g is t h e homomorphism induced by defi-ning g on the generators of the domain as follows: 2i g (a, A ) c o n t a i n s t h e term (a, A ) ; a c V th 2 i i . If A -+ abl ... b is the i production n of P a n d a E V then g ( a , A ) contains 1 2 i i i .There are no o t h e r terns in ga(a,L) .Note t h a t because g i s a hombmorphism, g ( A )= 4, w h e r e .?. * * is t h e identity of the monoid (X I ) T h e function 6 i s the canonical h o m o m o r p h i s m w h ' i c h * * coalesces a product i n (C T ) into a single ordered pair by component-wi se mcltiplicati3n of t h e first entries ( t h u s allowing cancellation in H) and catenation of the second entries. For e x a m p l e , If a c V p and A -+ ab ... b is t h e j th 1 r~ production of P then p contains the summand 3. p contains no other summands.k We adopt the convention that p = A f o r k ' 0. k Note that since p contains X, p contains A as w e l l as a l l summands of pJ f o r j ' k. For notational convenience w e adopt the followiag conventions. * * First; where no ambiguity can r e s u l t , products i n R(T: T ) of the form will be abbreviated as: occur in R ( c I ) . Second, we d e f i n e the function 'Yk as follows: where ai E V and p i s the polynpmial d e f i n e d above. Note that. if k < 0 , then y ( a a ... an) = v(ala2 ,.. a ) and Y~( A ) = A . notation, we can re-state the theorem as follows: Theorem (version 2) -: Let C = <VN, v~ ' P, S > b e a proper context-free grammar. Then there exist ms2s Y, g and 6 such t h i t + such that for every x E V x = xlxZ ... t a i n s a t e r m ( S , A ) if'and only if S ---X. T h e proof o f t h e theorem rests on t h r e e lemmas. -Lernma I i m p l i e l ; t h e \" i f \" p a r t of t h e theorem; Lemma 111 i m p l i e s the \"only i f \" p a r t . Lemma 1 1 i s used in the proof of Lemma 111.", |
| "uris": null |
| }, |
| "FIGREF28": { |
| "num": null, |
| "type_str": "figure", |
| "text": ", and A -M. Then f o r a l l k ' h ( A ) , k 6 g \\Ykcm) c o n t a i n s ( A , A ) . P r o o f (by induction on h(A), t h e h e i g h t of t h e d e r i v a t i o n A ) : k B a s i s : If h(A) = 0 , t h e n A = A and f l = A . Then Y~( A ) = p (A,Ei). Since A is a summand of p, it follows that (A,A) is a summand of k k p ( A , A ) , and t h e r e f o r e A , i s a summand of 6 g B k ( A , h ) . Thus t h e A k d e r i v a t i o n A A is r e p r e s e n t e d in 6 g Y (A) by ( A , A ) , which k e s t a b l i s h e s the b a s i s . A I n d u c t i o n : L e t A be a derivation of height n + 1, A -Y . By a s s e r t i o n 3, where and where h(ri) n. k t h e n by the i n d u c t i o n hypothesis, b g Y (M ) c o n t a i n s the summand k j k a j , r j ) C o n s i d e r t h e t e r m of g ' Y (M ) w h i c h c a n c e l s t o (a I? ) ' I T ) . T h i s t e r m must b e of t h e form ( a I ' )T, where r is 1' 1 1 E i t h e r a e V o r a c V p . T h e sum 6g k + l y a p r e f i x o f r (M1), which contains 6 g ( a 1' I' 1 )T. If al c V1, then g(al ,rl) .contains (Aa2ag: . . a , j rl) , and Gg(al , r l ) T c b n t a i n s r (ha2a3.. .a , j r l ) . O n the o t h e r hand, t h e sum bg k+lg r (M ) a l s o k+l 1 contains dpgny (M ) If a E V q , then (Aala2.. . a , j ) i s a s u m a n d l o w s t h a t 6g k+l (M) c o n t a i n s T h i s c o m p l e t e s t h e p r o o f .", |
| "uris": null |
| }, |
| "FIGREF29": { |
| "num": null, |
| "type_str": "figure", |
| "text": "L e t a E V , I' f 1 . For k 2 8 , all terms of g (a,l\") a r e of the form (b,aJ') (S ,A). . . (GI , A ) where b c V, c e 7 , m 2 0 , m i For n o t a t i o n a l . convenience w e abbreviate c c by N; Ilence we 1 m d e n o t e (b ,~r ) (c , A ) . . . (cl,A) by ( b N , A T ) . m Proof by induction on k , the number of applications of g . By 0 definition, g (a,I') = ( I ) w h i c h e s t a b l i s h e s t h e a s s e r t i o n f o r the v a l u e k = 0 . n+l Assume t h e a s s e r t i o n h o l d s f o r k < n and c o n s i d e r g a , = gq ( a , ? ' ) . By the i n d u c t i o n h y p o t h e s i s , a l l terms of $(a,T') a b f i ,~~' ) where b aN. Hence terms of g n f l ( a , r ) are o f the form g(bE,Or). S i n c e g l i m i t e d t o is t h e identity. . g ( b i ,~~) = [g(b,bT ) ] ( i , ,~) . By d e f i n i t i o n of g , g(b ,Or ) c o n t a i n s only terms of the i'orm (cG, j 9 1 ) j n't-1 w h e r e C + blf i s a p r o d u c t i o n . T h e r e f o r e t e r m s of g n c e C -h b?l and b =s, aN i t f o l l o w s that C aNM. k c o r o l l a r y : A l l terms of g (&r) a r e of t h e form (~N M , A T ) . k Lemma 111: I f 6 g , yk(M) c o n t a i n s ( A~, A ) , t h e n A -MN. Proof by--induction on the l e n g t h of M: Basis : L e t a. E V and assume k 6g Yk(a) c o n t a i n s (G>A). If pi r e p r e s e n t s a n a r b i t r a r y summand of p other t h a n PL, then every k t e r m of g Y (a) can b e r e p r e s e n t e d i n t h e form k where 0 r n < k and n denotes t h e number of n o n t r i v i a l summands of p which are f a c t o r s of t h e term. By c o n s t~u c t i o n , every summand of p is e i t h e r A -o r of the form + ( B .F , j i) where Bi I : VN, P c V , j i F -T 11. every term of g ( B . , j i) is of t h e f o r m : M.P , I ' . j ..) where Ci Vi, Mi, P c V , T i t l e same lemma, it follows that every term of g (a,A) is of the form c e every t e r m of g Y-(a) is of the form k ----\" P . M . for 1 r i r n and C is a tern1 t of g Bk(a) such t h a t 6 [ k ] = ( A~, A ) ; t must be in the form indicated above. In o r d e r f o r t to cancel under 4 , the following must be true: C1 = A since C cannot cancel from t, 1 --P = Q C for 1 i I n since C 2 -acn+l must all cancel from t. contains (AN,A) t h e n A =-. MN. L e t fi = Ma be a string k5 such t h a t I~a i = n+l and 6g y (Ma) c o n t a k s (AN,&). Because 6 g k and Y are hmornorphisms, k Then 6 g W (K) must c o n t a i n a t e r m (T A ) and 6g Y (a) must c o n t a i n k i s 1 k a t e r m (T A ) such t h a t T T = AN and h = r d e r f 6 r t h i s tc o c c u r , T2 must be of tahe form (BE*) r~h p r e * --B c ( V , N2 r V , and TI j u s t , b e of the form (ANIB) where1 A E V, -* I -N1 E V , and N = fi1i2. ( I f T and T w e r e n o t of t h i s form, 1 2 k c a n c e l l a t i o n t o ~ would be i m p o s s i b l e . ) Thus 6 g Yk(M) c o n t a i n s -(AN R,Al)., and by t h e induction h y p o t h e s i s 1 k A l s o 6 g Y (a) c o n t a i n s ( B & > , A~) and by t h e b a s i s k It f o l l o w s that and s i n c e A = M a and N = N2N1' which completes-t h e p r o o f . The theorem now f o l l o w s from Lemmas 1 and I I 1 and A s~e r t i o n 2. The ' i f ' p a r t f o l l o w s from Lemma I and A s s e r t i o r i 2 , and t h e 'only i f ' part f o l l o w s kminediately from Lemma I11 f o r the s p e c i a l c a s e of N = A . A s w e have s t a t e d the theorem, the l e n g t h of x i s used t o determine a s u f f i c i e n t number of a p p l i c a t i o n s of g and Y . A l t e r n a t i v e l y , the theorem could be foxmulated i n t e r m s of the heights of d e r i v a t i o n s of X; if A is a derivation of x of h e i g h t k, t h e n f o r e v e r y n 2 k, t h e t e r m ( S , A ) w i l l b e i n the p o l y n o m i a l . s~\" Y (x) . F u r t h e r m o r e , i t n f o l l o w s f r o m L e m m a 111 t h a t no harm i s done by c h o o s i n g t h e value of n too large, i -e . , no 'false' d e r i v a t i o n t e r m s will o c c u r . In t h e f l r s t statement of t h e t h e o r e m , the d e r i v a t i o n t e r m s n n n are o b t a i n e d from the polynomial B g Tl,p v ( x . ) which can be rewritten in the f o r m A l t h o u g h w e have used a c o n s t a n t v a l u e o f n ( e q u a l t o the l e n g t h of X) f o r both t h e powers o f the map g and the p o l y n o m i a l p , some economy can be g a i n e d in t h i s respect. In fact, the poweYs 5f g and p c a n decrease f r o m left t o r i g h t s o long-as they remain large enough t o p e r f o r m the a p p r o p r i a t e c o m p u t a t i o n s o n t h e suffix s t r i r l g s of X. Thus, the theorem is t r u e ( b~t c o n s i d e r a b l y mora difficult to prove) i f , o n e i n s t e a d uses a p a r s i n g p o l y n o m i a l o f t h e form V.", |
| "uris": null |
| }, |
| "FIGREF30": { |
| "num": null, |
| "type_str": "figure", |
| "text": "number of i n t d e s t i n g special cases occur based cln t h e c h o i c e of V1 and V 2 ' C a s e 1. V1 = VT.The f u n c t i o n g handles all p r o d u c t i o n s of t h e form w h i l e p handles productions of t h e f o r m N o t i c e that since g i s n o n t r i v i a l on o n l y V g need b e u s e d only T ' once; i-e., The parsing p o l y n o m i a l i s t h e n The s p e c i a l c a s e 01 V = VT and,V2 = VN 1 r e s u l t s in a particularly simple form if t h e grammar is in Greibach no m a 1 f o r m . The polynomial p = ( A , A ) and therefore has 110 effect. Since g need o n l y b e applied once, a l l derivations a r e found i n one s t e p . E x a m p l e 1: G = <~, A , B > , {-a,b), S , For t h e s t r i n g x = aabb, t h e p a r s i n g polynomial g [ Y (x)] t h e n c o n t a i n s k ( a m o n g other t h i n g s ) for a l l k 2 2, T h i s c o n t a i n s : [w(S,l) ( h , h ) ] [ ( A , 2 ) (:,A) ( x .~) ( A , a ) ( i ,~) ( x ,~) ] [ ( A", |
| "uris": null |
| }, |
| "FIGREF31": { |
| "num": null, |
| "type_str": "figure", |
| "text": "( s , I . ) ( H , A > ] [~ ( A , 3 ) ] [ ( B , 4 ) ] [ ( B , 4 ) ] a f t e r one a p p l i c a t i o n of g , [(S,l)(A,~>][(A,223)(B,h)(B,h)] [ ( B , 4 ) ] [ ( , 4 a f t e r t h r e e . Applying 8 r e s u l t s i n ( S ,122344) as b e f o r e . C a s e 3 . Vl = 0. N. ow the e n t i r e p a r s e i s h a n d l e d bp p . The p a r s i n g polynomial becomes V I . . O b s e r v a t i o n s +he m a j o r theorem p r e s e n t e d here shows how c o n t e x t -f r e e p a r s i n g may b e c a r r i e d out by p u r e l y a l g e b r a i c means. A l l parses", |
| "uris": null |
| }, |
| "FIGREF32": { |
| "num": null, |
| "type_str": "figure", |
| "text": "number of terms of a p a r s i n g p o l y n o m i a l f o r a s t r i n g x c V is T u n r e a s o n a b l y l a r g e . %lowever, m o s t of the t e r m s i n such a p o l y n o m i a l are n o t a s s o c i a t e d w i t h a d e r i v a t i o n i n t h e grammar, and method; e x i s t f o r r 9 d u c i n g t h e c o m p u t a t i o n by d i s r e g a r d i n 4 dead-end t e r m s before t h e y a r e c o m p l e t e l y e v a l u a t e d . By a p p l y i n g s u c h t e c h n i q u e s i n a s t r a i g h t f o r w a r d f a s h i o n , and c h o o s i n g V and V 2 i n v a r i o u s ways, 1 l y z e r . C a s e 2 i s the a l g e b r a i c version of generalized bcttorn-up. P a r s i n g algorithms are typically so d i f g erent one from a n o t h e r that they are incomparable. B u t u s i n g techniques d e s c r i b e d above, m a n y parsing algorithms m a y b e p o s e d in a single a l g e b r a i c f r a m e w o r k . T h i s m a y f a c i l i t a t e t h e comparison and evaluation of p a r s e r s and of various c l a s s e s of g r a m m a r s .", |
| "uris": null |
| }, |
| "FIGREF33": { |
| "num": null, |
| "type_str": "figure", |
| "text": "first the automatic identification of linguistic entities useful for the representation of document content, and then the assignment to the prospective content identifiers of weights reflecting their importance for content description. Since these tasks must ultimately depend on a study of the texts or documents under consideration-, a g r e l t deal can be learned by examining t h e occurrence p a t t e r n s of words and o t h e r l i n g u i s t i c e n t i t i e s i n t h e documents of a c o l l e c t i o n . Indeed, among t h e t h e o r i e s of term importance which havebeen s t u d i e d i n r e c e n t y e a r s , the b e s t known 'ones a r e based on t h e r e s p e c t i v e frequency d i s t r i b u t i o n s a c r o s s a v a r i e t y of w r i t t e n t e x t s . A ) Variance-Based Measures The most widely used of the s t a t i s t i c a l t h e o r i e s ' d i s t i n g u i s h e s so-called \" s p e c i a l t y \" wowds from \"nonspecialtyll words by assuming t h a t a d e v i a t i o n from randomiiess i n t h e occurrence p a t t e r n of c e r t a i n t e x t words i s i n d i c a t i v e of s p e c i a l i z a t i o n and hence of good content i d e n t i f i e r s . Thus t h e best content d e s c r i p t o r s are t e r m s Whose occurrence pattepns deviate most s t r o n g l y from randomness. Since a random sprinkxing of t h e occurrences of a given text w~d a c r o s s t h e documents o f a c o l l e c t i o n l e a d s t o wora frequency d i s t r i b u t i o n s which follow t h e Poisson model, a compa~ison o f t h e actual freqiieiicy c h a r a c t e r i s t i c s o f a given term with t h e Poisson d i s t r i b u t i o n l e a d s t o t h e a p p r o p r i a t e d i s t i n c t ion between good content words and poor. ones. More specifically, since the variance vk o f t h e frequency d i s t r i b u t i o n o f term k i s p r o p o~t i o e a l t o t h e t o t a l frequency of occurrence F~ f o r t e r m s whose d i s t r i b u t i o n obeys t h e Poisson model, a measure of term importance i s k o b t a i n a b l e by using a formula based on t h e r a t i o of vk t o F . Some t y p i c a l formulas used f c r t h i s purpose a r e vk/fk and n k k 2 -v / F where n is t h e c o l l e c t i o n s i z e . er of 22c.~-;::r I.. --------.., I li.ec_uency of term k in d o c m e n t l I binary f r e~u e n c y of T e r m k i n document i t o t a l frequency of term k in collection document Srequency of term k in collect ion (number d f documents i n which t h e term occurs ) average frequency of term k in Term Val ue Measurements One such variance-based measure used by Dennis under t h e name of NOCC/EK [ 3 ] may be computed as It is obvious from this formulation t h a t t h e most e f f e c t i v e terms are those whose occurrence frequencies f k i n the i n d i v i d u a l documents d e v i a t e strongly i k from t h e average frequency F /n. B) Signal-Noise Measure Another measwe based on t h e c h a r a c t e r i s t i c s o f t h e frequency d i g t r i b u t i o n of individual text u n i t s across t h e documents of a c o l l e c t i o n is t h e signal-noise r a t i o which varies w i t h t h e skewness of t h e frequency d i s t r i b u t i o n . This measure has t h e form O F entropy and a s s i g n s t h e highest value ts those terms whose occurrence c h a r a c t e r i s t i c s e x h i b i t t h e g r e a t e s t v a r i a t i o n from one document t o another; c c n t r a r i w i s e low values are assigned t o terms with r e l a t i v e l y similar frequency p a t t e r n s i n each o f t h e documents of a c o l l e c t i o n . [3,4] The idea is t h a t terms with even frequency d i s t r i b u t i o n s which may occur an i d e n t i c a l number of times i n each document of t h e c o l l e c t i o n canr~ot be used t o d i s t i n g u i s h t h e documents from each o t h e r ; hence, their assignment f o r purposes of content r e p r e s e n t a t i o n is counterproductive. The r e v e r s e o b t a i n s f o r terms with skewed fvequency d i s t r i b u t i o n s .", |
| "uris": null |
| }, |
| "FIGREF34": { |
| "num": null, |
| "type_str": "figure", |
| "text": "The signal n o i s e value (S/N) f o r term k i s defined a s k A 1 T e r m V a l uc Meas urcmcnt s k The negative term i n expression ( 2 ) i s known.as t h e n o i s e N ; it is k k maximized f o r even d i s t r i b u t i o n s where fk = F /n f o r a l l f . . The 1 p r o p e r t i e s of t h e s i g n a l -n o i s e measure are t h u s very similar t o t h o s e described e a r l i e r f o r t h e variance-based formulas. C) I nf ormat ion Theoret ic Considerations The f o r~g o i n g development l e a d s t o a d i s t i n c t i o n among t h e terms i n k accordance w i t h t h e r e l a t i v e sizes of t h e i n d i b i d u a l t e r m f r e q u e n c i e s fi k i n t h e documents and t h e t o t a l c o l l e c t i o n frequency F . A q u e s t i o n arises about t h e p r e f e r r e d s i z e of the c o l l e c t i o n frequency F~ (or of t h e k document frbquency B 1 f o r terms that are u s e f u l as c o n t e n t i d e n t i f i e r s .This problem may be t a c k l e d by having r e c o u r s e t o c e r t a i n i n f o r m a t i o n -t h e o r e t i c concepts. Consider t h e t a s k o f supplementing a s e t of e x i s t i n g 5ndex t e r m s i d e n e i f y i n g a c o l l e c t i o n o f documents by a d d i t i o n o f a c e r t a i n number o f newT e r m s . Each new ' t e r m i s t h e n most e f f e c t i v e when a) it provides maximum a d d i t i o n a l r e d u c t i o n i n u n c e r t a i n t y amongt h e documents of t h e c o l l e c t i o n ( t h a t i s , i t s assignment breaks up e x i s t i n g s u b s e t s o f documents that cannot be d i s t i n g u i s h e d by the e x i s t i n g term assignments i n t o s u b s t a n t i a l l y s m a l l e r s u b s e t s ) ; b) it e x h i b i t s l i t t l e redundancy w i t h t h e previously a v a i l a b l e terms SO t h a t i t s assignment does indeed optimally d i v i d e t h e v a r i o u s document s e t s .", |
| "uris": null |
| }, |
| "FIGREF35": { |
| "num": null, |
| "type_str": "figure", |
| "text": "", |
| "uris": null |
| }, |
| "FIGREF36": { |
| "num": null, |
| "type_str": "figure", |
| "text": "document frequency B , t h a t i s , t h o s e a s s i g n e d t o very f e w documents i n t h e c o l l e c t i o n , because their assignment provides l i t t l e a d d i t i o n a l d i s c r i m i n a t i o n among t h e documents; t h e second p r o p e r t y , on t h e o t h e r hand, does not o b t a i n for terms of high document frequency t h a t may be assigned t o a very large number of documents, because such terms w i l l obviously e x h i b i t a good d e a l of redundancy with the already e x i s t i n g terms.", |
| "uris": null |
| }, |
| "FIGREF37": { |
| "num": null, |
| "type_str": "figure", |
| "text": "g e r thanand f o r some o t h e r s fi i s much smaller than -. n n D ) The Discrimination Value Model The discrimination value model uses a s a p o i n t of departure t h e r e t r i e v a l c a p a b i l i t y o f t h e various index terms; s p e c i f i c a l l y , a good content-indicative term is designed t o help i n t h e r e t r i e v a l of m a t e r i a l t h a t i s wanted ( t h u s enhancing t h e r e c a l l ) , and i n the r e j e c t i o n of m a t e r i a l t h a t is extraneous (thus enhancing the precision)fi. To produce h i g h r e c a l l , t h a t i s t o retrSeve most everything t h a t i s r e l e v a n t , t h e terms used t o 'identify documents and u s e r queries must be fairly general in n a t w e ; high p r e c i s i o n , on the o t h e r hand, t h a t is t h e r e j e c t i o n of t h e nonreleudat m a t e r i a l , depends on t h e use of reasonably s p e c i f i c content i d e n t i f i e r s . The indexing problem then reduces t o t h e choice of terms t h a t are s p e c i f i c enough t o prohuce h i g h p r e c i s i o n while also being general enough t o produce high recall. I n the discriminatiqn value model, t h e assumption is aade t h a t t h e b e s t terms in t h i s r e s p e c t a r c those which cause t h e maximum p o s s i b l e separation among t h e dobuments i n t h e \"document space\". Consider , i n p a r t i d u l a r , a c o l l e c t ion of documents each i d e n t i f i e d by a s e t of content i d e n t i f i e r s , o r index terms. The ?'ndex term sets for two given documents can be compared t o produck a s i m i l u i t y c o e f f i c i e n t measuring the closeness between t h e r e s p e c t i v e documents. * Recall is t h e proportion of r e l e v a n t m a t e r i a l r e t r i e v e d while precision is t h e proportion of r e t r i e v e d material t h a t i s r e l e v a n t . An e f f e c t i v e -7 r e t r i e v a l system is one whlch produces t h e h i g h e s t possible p r e c i s i o n f o r a given l e v e l of r e c a l l . T e r m V a l uc> Mcas urcmcnts The e x i s t e n c e of t h e term q e t s r e p r e s e n t i n g t h e v a r i o u s documents, and t h e p o s s i b i l i t y of computing s i m i l a r i t y measures between documents can be used t o d e f i n e a document space For t h e c o l l e c t i o h . I n such a space two documents appear i n c l o s e proximity when t h e i r s i m i l a r i t y a o e f f i~i e n t i s l a r g e ; c o n t r a r i w i s e , documents e x h i b i t i n g l i t t l e s i m i l a r i t y a r e widely separated i n t h e document space. One may t h e n c o n j e c t u r e t h a t a document space which is \"bunched up\", i n t h e s e n s e t h a t a l l documents e x h i b i t somewhat similar term sets i s not u~e f u l f o r r e t r i e v a l , s i n c e one document cannr*t t h e n be d i s t i n g u i s h e d *om another. On t h e con;trary, a space. Which is spread out i n s u c h m a way t h a t t h e documents a r e widely s e p a r a t e d from each o t h e r may provide a n i d e a l r e t r i e v a l s i t u a t i o n s i n c e some documents may t h e n be r e t r i e v e dhopefully t h e r e l e v a n t oneswhile o t h e r s can be r e j e c t e d . This suggests that t h e v a l u e of an index term can b e a s c e r t a i n e d , b y measuring t h e amount o f spreading i n t h e document space which occurs when that term i s assigned t o t h e documents o f t h e c o l l e c t i o n . S p e c i f i c a l l y ; i f Q i s t h e d e n s i t y o f t h e document space without term k p r e s e n t among t h e content i n d i c a t o r s , and Qk i s t h e d e n s i t y a f t e r term k i s assigned, then f o r a good t e r m Q -Qk > 0, s i n c e t h e space w i l l have spread a f t e r term k i s assigned. ConverseSy f o r poor terms Q -% T 0.2 [5,6] An a p p r o p r i a t e The d e n s i t y o f t h e space might be computed, f o r example, as t h e sum of a l l pa:-vwise s i m i l a r i t i e s between d i s t i n c t document p a i r s , t h a t is where S(Di, D.), 0 < S < 1, i s t h e s i m i l a r i t y between documents D 3 -and D.. i measure of term importance i s then t h e term discrimination value, DVk 3 defined as It may be of i n t e r e s t t o i n q u i r e i n t o the r e l a t i o n s h i p between t h e discrimination value of a term and the s t a t i s t i c a l . (frequency) parameters introduced e a r l i e r . The following conclusions are reached from a study of t h e indexing vocabularies i n s e v e r a l different s u b j e c t areas, r e l a t i n g t h e document frequency of a term t o i t s discrimina.tion value: [ 5 ] a ) terms with yery Low documeht fiequenay t h a t may be assigned t o very feQ documents i n a c o l l e c t i o n a r e generally poor discriminators; when t h e terms are arranged i n decreasing order of their discriminamtion values (where rank 1 i s a s d g n e d t o the b e s t discriminator, rank 2 t o the next b e s t , and s o on) such terms e x h i b i t ranks i n excess of t / 2 for a t o t a l of t e x i s t i n g terms; b) term3 with high document frequencies, comprising those t h a t are assigned t o more than 1 0 percent of t h e documents of a c o l l e c t i o n a r e t h e worst discriminators, with average discrimination ranks (ranks i n decreasing discriminatioh value order) near t ; c) t h e b e s t discriminators are those whose document frequency is n e i t h e r \u20acQO high nor too low -with document frequencies between n/100 and n/10 for n documentq; t h e i r average discrimination ranks are generally belaw t / 5 f o r t terms. The vector space a n a l y s i s then appear& t o confirm t h e conclusions derived earlier from t h e s t a t i s t i c a l models, t h a t terms which appear i n a c o l l e c t i o n with great r a r i t y or excessive frequency are not optimal for content description purposes. Term V a l u e Measuremt:nts 2. Compariscn and Evaluation The d i s c r i m i n a t i o n value a n a l y s i s can be used t o d e r i v e an e f f e c t i v e indexing p o l i c y : s i n c e t h e b e s t terms appear t o be t h o s e w i t h medium document f r e q u e n c i e s , such terms can be d i r e c t l y assigned a s c o n t e n t i d e n t i f i e r s without f u r t h e r r e f i n i n g t r a n s f o r m a t i o n s . On t h e o t h e r hand, terms with excessively high document f r e q u e n c i e s must be made more s p e c i f i c t h e r e b y decreasing t h e frequency of t h e i r assignment t o The q u e r i e s and documents", |
| "uris": null |
| }, |
| "FIGREF38": { |
| "num": null, |
| "type_str": "figure", |
| "text": "component, and vice-versa f o r a term class which r e p l a c e s a number o f i n d i v i d u a l class elements. It was shown e a r l i e r t h a t t h e use of p h r a s e s and term c l a s s e s ( t h e s a u r u s ) constructed i n accordance w i t h t*he frequency requirements imposed by t h e d i s c r i m i n a t i o n value t h e o r y produces s u b s t a n t i a l improvements i n r e t r i e v a l e f f e c t i v e n e s s ( r e c a l l and p r e c i s i o n ) . I n t h e p r e s e n t work, a d d i t i o n a l r e l a t i o n s h i p s a r e examined between t h e s t a t i s t i c d and t h e v e c t o r space models.However, i n s t e a d o f a o t u d l y 'using t h e v a r i o u s term s e t s i n a r e t r i e v a l environment, an attempt is made t o r e l a t e t h e formal frequency and v e c t o r spaee p r o p e r t i e s of t h e terms t o t h e se-nantic c h a r a c t e r i s t i c s o f t h e s e terms. S p e c i f i c a l l y , c o n s i d e r a c o l l e c t i o n o f documents i n a given s u b j e c t a e a and an a p p r o p r i a t e set of u s e r q u e r i e s p e r t a i n i n g t o t h a t a r e a . For each u s e rquery, t h e set of documents can be p a r t i t i o n e d i n t o two s u b s e t s c o n s i s t i n g o f t h e r e l e v a n t s e t R and t h e rlonrelevant s e t I , r e s p e c t i v e l y . Relevance i s assumed t o be u s e r -s p e c i f i e d i n s u c h a way t h a t a r e l e v a n t itemi s assumed t o be one which i g r e l a t e d i n some s e n s e t o t h e i n f o r n a t i o n need expressed by t h e v a r i o u s u s e r queries. The l i n g u i s t i c , o r s e m a n t i c , c h a r a c t e r o f a given term can now be i n t r o d u c e d by assuming t h a t t h e most v a l u a b l e contenti d e n t i f i e r s a s s i g n e d t o a c o l l e c t i o~l o f t e x t s are t h o s e which are! c o n c e n t r a t e d i n t h e documents s p e c i f i e d as r e l e v a n t t o t h e respective q u e r i e s , as opposed to the. nonrelevant ones, c o n t r a r i w i s e , t h e l e s s v a l u a b l e t e r w w i l l be concentbated i n the n o n r e l e v a n t items. The d i s c u s s i o n may b e f o r m a l i z e d by u s i n g t h e concept o f t e r m r e l e v a n c e TR.[ 7 ]Consider a t e r m k c o n t a i n e d i n query Q;. t h e terBm r e l e v a~k e TR(k) may be d e f i n e d as", |
| "uris": null |
| }, |
| "FIGREF39": { |
| "num": null, |
| "type_str": "figure", |
| "text": "average o f t h e r e l e v a n c e v a l u e s o b t a i n e d f o r t h e v a r i o u s queries. The mathernaticxlly u n d e s i r a b l e s i t u a t i o n when I RI r o r when h 0 is n o t l i k e l y t o o c c u r i n a p i -a c t i z d envircnmegt. k k T e r m V a l u e Mcas urcmen ts It i s c l e a r from the function ( 4 ) t h a t high values arc ~: S S~~T I C~ t o", |
| "uris": null |
| }, |
| "FIGREF40": { |
| "num": null, |
| "type_str": "figure", |
| "text": "v e r i f y th'e r e l a t i o n s h i p s between the s t a t i s t i c a l models of word importance and t h s vector space model, dcsument c o l l e c t i o n s a r e used i n three d i f f e r e n t s u b j e c t a r e a s , i n c l u d i n g aerodynamics (cRAN), medicine (MED) and world a f f a i r s (TIME). The vocabularies and u s e r populations are d i s j o i n t f o r t h e s e r h r e e a r e a s . R e s u l t s which c a r r y through for a l l t h r e e cases should be extendable t o o t h e r subject f i e l d s a s w e l l . The basic c o l l e c t i b n s t a t i s t i c s are contained i n Table 2. It may be seen from the Table t h a t t h e t e r m r e l e v a n c e i s d e f i n e d for only a r e l a t i v e l y small number of terms f o r each c o l l e c t i o n , namely 458, 1 7 2 and 375 for CRAN, MED, and TIME, r e s p e c t i v e l y . The r e a s o n Ls t h a t a term relevance value is computable only f o r terms which occur j o i n a l y in certain query-document p a i r s . Fop small experimental c o l l e c t i o n s o p e r q t i n g with a r e s t r i c t e d number of q u e r i e s t h e s i z e of the corresponding term sets i s obviously l i m i t e d . Consider now the comparison o f t h e standard s t a t i s t i c a l term value measures with t h e t e r m d i s c r i m i n a t i o n v a l u e s obtained by t h e v e c t o r space transformations. Table 3 shows t h e v a l u e s of t h e NOCC/EK and S/N measures (expressions (1) and ( 2 ) ) obtained f o r tine 50 terms w i t h h i g h e s t d i s c r i m i n a t i o n values and t h e 50 terms with lowest d i s c r i m i n a t i o n v a l u e s f o r each o f t h e t h r e e test c o l l e c t i o n s . The range of t h e r e s p e c t i v e values is given i n each case, as well as t h e average values f o r each s e t of 50 terms i n p e r c e n t ( t h a t i s , on Term Value Mcdsurcmcnts", |
| "uris": null |
| }, |
| "FIGREF41": { |
| "num": null, |
| "type_str": "figure", |
| "text": "Number of t e p s occurring j o i n t l y in queries and document sets 0 to 100). T test values are.also shown ~e p r e s e r l t i n g t h e p r o b a b i l i t y t h a t t h e two sets of 50 values (for the high DV and low DV terms) could have been d e r i v e d from a common probability d i s t r i h u t z o n by chance. In statistical significance testing, a t -t e s t value smaller than 0.05 i s normally taken t o imply a s i g n i f i c a n t difference; that is, the hypothesis that t h e mo s e t s of values do i n fact o r i g i n a t e from a common d i s t r i b u t i o n is r e j e c t e d in such a case. [8] It rpay be seen that the ranges of values for the s t a t i s t i c a l parameters NOCC/EK and S/?$ e x h i b i t s u b s t a n t i a l differences for a i l three c o l l e o t i o n s . The same i s true for the corresponding average values. Moreover t h e d i f f e r e n c e s are in a l l cases statistically s i g n i f i c a n t . , It i s then clear t h a t a high d i s c r i m i n a t i o n value r e f l e c t e d in the a b i l i t y of a term t o expand t h e document space upon assignment t o t h e c o l l e c t i o n also implies, favorable statistical parameters i n terms of v a i a n c e and skewed frequency distributions; t h e converse is t r u e for t h e low d i s c r i m i n a t i o n values.", |
| "uris": null |
| }, |
| "FIGREF42": { |
| "num": null, |
| "type_str": "figure", |
| "text": "are t h e m u l t i p l i c a t i v e factors which relate t h e average values f o r the 50 high discriminators and t h e 50 low discrimihators for each of t h e three measures ( t h a t i s , t h e factor by Term V a l u e Measurements which t h e OW average value must be m u l t i p l i e d t o obtain t h e h i g h ) . I t may be seen t h a t t h i s f a c t o r is much higher f o r t h e term relevance than f o r e i t h e r of NQCCIEK o r S/N. The a c t u a l f a c t o r s f o r t h e term relevance are 6.66, 80.0 and 36.33 f o r t h e CRAN, MED, and TIME c o l l e c t i o n s , r e s p e c t i v e l y . Thi& i n d i c a t e s t h a t t h e high d i s c r i m i n a t o r s have very much higher average term relevance t h a n t h e low d i s c r i m i n a t o r s ; a l t e r n a t i v e l y expressed, t h e r e i s s u b s t a n t i a l agreement between t h e semantic term relevance concept and t h e automatically derived term d i s c r i m i n a t i o n values.", |
| "uris": null |
| }, |
| "FIGREF43": { |
| "num": null, |
| "type_str": "figure", |
| "text": "contains range and average v a l u e s f o r NOCC/EK, S/N, and DV f o r t h e 50 terms with highest term p r e c i s i o n and t h e 50 terms with lowest p r e c i s i o n f o r t h e CRAN and TIME c o l l e c t i o n s , r e s p e c t i v e l y . Averages a r e produced f o r only 30 high and 30 low p r e c i s i o n terms f o r t h e MED c o l l e c t i o n because i n t h e medical environment t h e small number of available q u e r i e s ( 2 4 ) made it p o s s i b l e t o compute term p r e c i s i o n values f o r only 172 terms i n a l l .It i s c l e a r from t h e output of Table 4 t h a t t h e d i f f e r e n c e s i n t h er e s p e c t i v e values aye s u b s t a n t i a l i n a l l cases, and t h e t -t e s t values i n d i c a t e t h a t they arc f u l l y s i g n i f i c a n t .For t h e t h r e e c o l l e c t i o n s under study, &he evidence i n d i c a t e s t h a t terms with favorable formal parameters tend t o be concentrated i n documents i d e n t i f i e d as r e l e v a n t by t h e u s e r population, and vice-versa f o r terms with unfavorable formal parameters. Also shown i n 4 -k Table 4 are average document frequency ( B ) and average t o t a l frequency (F ) values f o r t h o high and low relevance terms respectively. I t may be seen t h a t t h e T e r m V a l u e Mcas urements h i g h relevance terms e x h i b i t a much lower frequency spectrum (as e~p e c t e d f o r good d i s c r i m i n a t o r s ) than t h e l o w r e l e v a n c e terms. Once again, i t appears t h a t t h e term relevance r e f l e c t i n g t h e semantic p r o p e r t i e s of t h e terms i n t h e i r p a r t i c u l a r c o l l e c t i o n environment e f f e c t s a d i v i s i o n among t h e terms very si~ilar t o t h a t obtained by t h e d i s c r i m i n a t i o n value cornputat ions. I n e a r l i e r work it w a s shown t h a t t h e d i s c r i m i n a t i o n value theory which leads t o t h e assignment t o queries and documents of medium frequency terms cincluding a l s o phrases constructed from high frequency terms, and term classes made up of low frequency terms-) e x h i b i t s e g f e c t i v e r e t r i e v a l c h a r a c t e r i s t i c s . [4,5,6] Typical average r e t r i e v a l p r e c i s i o n values f o r t h r e e d i f f e r e n t r e c a l l l e v e l s ( r e c a l l of 0.1, 0.5, and 0.9) a r e shown f o r t h e three c o l l e c t i o n s i n Table 5. The output shows t h a t t h e use of mediumfrequency phrases and term classes improves performance by about 20 percent compared with the assignment of s i n g l e terms alone. The comparison o f Tables 3 and 4 between discpimination values on the one hand, and s t a t i s t i c a l and semantic parameters on th\"e o t h e r , i n d i c a t e s t h a t the same theory which produces such e f f e c t i v e r e t r i e v a l c h a r a c t e r i s t i c s a l s o conforms t o the known s t a ' t f s t i c a l and l i n g u i s t i c t h e o r i e s of t e r m behavior.", |
| "uris": null |
| }, |
| "FIGREF44": { |
| "num": null, |
| "type_str": "figure", |
| "text": "---------. I . ---------.-----------,,,--,------S/N range 2,792 t o 0.693 1.738 t o 0.126 average ( in percent ) 48: 46%", |
| "uris": null |
| }, |
| "FIGREF45": { |
| "num": null, |
| "type_str": "figure", |
| "text": ", --------------------. -----------------------Comparison of T e r m Relevance with Term D i s c r i m i n a t i o n Values", |
| "uris": null |
| }, |
| "FIGREF46": { |
| "num": null, |
| "type_str": "figure", |
| "text": "-----I ----------------1.953 to 0.000 0.998 to 0.045 average 42.81% average 20.63% t-test 0.00002 average high/average low 2.08---------------------------m Va l u e Measurements b,; MED 450 C o l l e c t i o n Comparison of Term Relcvarlce w i t h T e r m D i s c r i r n i n a t i o n Values ( cont . )", |
| "uris": null |
| }, |
| "FIGREF47": { |
| "num": null, |
| "type_str": "figure", |
| "text": "g h / a v e r a g l o w 1.11 -T e r m V a l u e Measurements c) TIME 425 Collection C~m p a r i~b f i of T e r n Relevance w i~h Term Discriminat i o n Values ( cont . ) -A --d -----------", |
| "uris": null |
| }, |
| "FIGREF48": { |
| "num": null, |
| "type_str": "figure", |
| "text": "", |
| "uris": null |
| }, |
| "FIGREF49": { |
| "num": null, |
| "type_str": "figure", |
| "text": "calls the function given by its argument and if s u c c e s s f u l pushes the structure returned by t h e function onto a s t a c k (SAVEQ) and a s s i g n s the structure t o the Q r e g i s t e r . sets the values of registers. It has three arguments level, register name, and value. Each cavil ~f SETR causes the r e g i s t e r name s p e c i f i e d to be placed on -3 list for the s p e c i f i e d level. SETR e n r r i e s are treated as stacks, providing automatic saves for recursi-ve calls. returns the contents of the register name specified by its argument, and pops it o f k the stack saving the last value. looks at the value of the register name specified by its argument without popping it off the stack. changes the vaLue of a register without changing s t a c k Levels. looks up the f e a t u r e v a l u e for a f e a t u r e speciFiied by i t s argurnebt of the current value of the word. register. Any word can be s p e c i f -t e d by giving a second aqpment. Tf GETF f a i l s for the word it looks at the root form of the w o r d for certzain features l o a~s up the word c l a s s o f the word specified by i t s argument. has as i t s argumcnt, t h e new s t a t e l a b e l . It p u s h e s the l a b e l o n t o a s t a c k (PATH) ; o u t p u t s t:he s-t:at-c? , o u~: p l l t s the contents o f t h e Q r c g J s t x r , and transfers c o n t r o l t.o t h e new s c a r e . backs to ehhe s t a t e s p e c i f i e d b y i t s nrgrmic~nt.", |
| "uris": null |
| }, |
| "FIGREF50": { |
| "num": null, |
| "type_str": "figure", |
| "text": "t s f o r t h e w o r d s p e c i f i e d b y its argument:. BUILDS b u i l d s a structure from trhe r e g i s t e r name l i s t : . SNOPAR leve'ls. In the examination s t a g e , traces may be turned on l e x i o a l entries m a y b~ examined or minor changes to the grammar may be made. Functions available A f o r the examination of stacks, registers and lexicon are POP, OUT, GETR, LOOKLEX,-and TRACE. A function GETENG is also available for dictionary lookup in other languages. PARSER requires approximately 150 lines of SNOBDL code and i s c u r r e n t l y operating on a DEC 10. A hatch version has been t e s t e d on an ZBM 360 I n order t o use PARSER, a g r a m m a r and lexicon must be developed a s disc: f i l e s . Since the grammar L s developed as a separate file d i f f e r e n t components of t h e grammar can be tested and p u t together i n a variety of c o n t i g u r a t i o n s . I f a Lextcon i s n o t developed as a d i s c Eiie p r i o r to a parse, it may be e n t e r e d fron the terminal A simple grammar which produces s u r f a c e structure t r e e s is shown in Example 1 along with a sample parsing. A p o r t i o n of t h e lt?xicon i s shown a t the bottom o f the p a g e . Example 2 shows the use of the GETF f u n ct i o n t o handle agreement between p l u r a l adjccti.vcs and a p l u r a l , r n a r k~r in Angas, a N i g e t i c l n lony,unjic. Epomplc '3 shows a gr;lm-111:lr w h i c h hyncll c?s r;c)ni-cnctt c?mbcdtli nf: i n Engl i :;h . fiornc nnmlr 1 cB i I T ; arcA :;hewn . 7 ' 1 1~ mc~dr~l I I : ; P~ r ihca E X : I I I I~ I c l '3 j ; r ; l i J l r I i ; i r j ! ; I , : , I :; i cij l 1 y I 11c o r l c A (Icvca 1 opctl i 11 En); 1 j :;I) 7'rl111;; f o r-tll;i 1 j O I I , J 1 ~I I : y , J o :I I I~ J < o : ; (~T I~J~J [ J~I . A J :; i ( (':I N V ~: J -: I I I I~I I ; I r f-o r 1 j 1 1 I ; wc: 1 1 ;I ;I. I i : or i c A 1 l l c~t l r r~ r for (:bo(* l ;JW (.in h~ic.ri c a n I n d i a n 1 nn):u;ij:cs) I i 11 tl(~vc.1 oplocJnI . SNOPAR The complete SNOPAR system has in a d d ' i t i o n to PARSER a routine for generating grammars from a .state transition graph and a register action table. This routine called NEW guides the user through a state transition graph and register actions to produce a grammar compatible with PARSER. Thd SNOPAR NEW routine is s t i l l in develoj~rnen~. The current routine allows deuelopment.of small grammars. The new developments will pro vide diagnostics of grammar errors. SNOPAR dlso has a line editor ( F I X U P ) and disc 1-10 commands. The complete system allows repetitive testing of model grammars, permits editing;and has trace capabilities fsr grammar debugging. * T E ? l S E c , G E T F ( 'TNS'))", |
| "uris": null |
| }, |
| "FIGREF51": { |
| "num": null, |
| "type_str": "figure", |
| "text": ",'PP',Q) 1 (TO( . T R Y V P P ) ) S E T R ( . V P , *NP',Q) I(TO(.POPVP)) V P = BUILDS(VP) [ R E T U R N ) TY L E X E N G . 1", |
| "uris": null |
| }, |
| "FIGREF52": { |
| "num": null, |
| "type_str": "figure", |
| "text": "( N ) ( N b S P L ) . . G I R L = (N)(tl9!{ L I N G ) . G L R L S -(E'0ItE.I I ) 1 P L ) . M A N = ( N ) ( t . l n r i S L l i G ) .M E N = ( N ) ( t ! U R P L ) .W O M A N = ( N ) ( N E R : ; L~I c ) .W O H E N = (N) (1132 P I , ) . T A B L E = ( N ) (!:!3R S f !lC) .", |
| "uris": null |
| }, |
| "FIGREF53": { |
| "num": null, |
| "type_str": "figure", |
| "text": "( i ! s r r~) ( E N G W O Y A N )' = ' ( A D J ) ( P L -P L l ) ( E N C G O O D ) ' L < * R~~' J ' -F f I J i T O > = ' ( A D J ) (P'L P L ) (EMG G O O D ) ' L < * B I J I M * > = [ A D J ) ( P L -PL) ( E N G N I f 3 ) ' =. ( I ) F T ) ( I . , N C ' I ' I l J~~)~ L < '~J A ' > = * ( D l~T ) ( b , t l~~ T t ! b ; )t c ' c~; ' , = ' ( D E T ) (E!;'; K I -) = ' ( K l . ) ( E t l C ; POS:;:.:;IVE) ' E X $ % STATE P O P N P COmYTLE!?ENT S T R I N G :", |
| "uris": null |
| }, |
| "FIGREF54": { |
| "num": null, |
| "type_str": "figure", |
| "text": "'SU!3J*p '(PRO Y O U ) * @ ) a[TO(,PQPS))\u20ac 4", |
| "uris": null |
| }, |
| "FIGREF55": { |
| "num": null, |
| "type_str": "figure", |
| "text": "= B U I L D S ( * / r J P / !~P P / P O S S . / @ S k T H ( , f i E r \" N P P C N r Q ) E V P P (~N~I e s T o ( , N p n > ) P E-0 GFTF('CASEP) b P O S * rS(TO(,POSPAO)) E F T u ( , f~P r 'PPOCpQ1 1 ( T Q ( r P Q P f J P 1 )B L N P CAT('ADJC) t F d T O ( p ? d B L ) l S E T P I ,!IF? 'ADJ' R ( , V P I C V P P o W I G )E U I . ' P ( ? N ' ) : (TO[ ,NTPP)'l S E f R ( ,VP, * A D J g r Q ) ' f E S T F ( * T Y P E o ) , 'C*) ISCTESTR(~AUX') @ B E e ) $ F ( T O [ , ' S R Y E S I ) S E ?~( , V F I~V '~G E Z F~' A Z J )", |
| "uris": null |
| }, |
| "FIGREF56": { |
| "num": null, |
| "type_str": "figure", |
| "text": "I S ( P < ' P O ' > ) t F ( T Q ( , P O P V P ) ) PAFSE(IO()) 8 S~~Q ( 8~I o L l > F ( T O [ , P O P V f a ) )", |
| "uris": null |
| }, |
| "FIGREF57": { |
| "num": null, |
| "type_str": "figure", |
| "text": "0U 11lStt4T T O E: :%PIItiE THE PEG1 STEF'? NO ItjPLIT Z-TFUSTUPE TO BE PRFIED I IlltitiT T O 150 1. IIIH~~T TD GO .:'Thi~ I.COPIFLEbIEIil ITF'I ti t::GO ElJIL-11 ITPI-IIZTI-~FE: -1 1 T'r'PE L . 1 I , UE 1 I.I~F-1-PF 0 1 1 1 I I FFED 1 $ ' 1 T i d PF'E' 4 1 I *; T l~lRt{T-& -I I { z I T '~' P E '~I C L I I -l-IE,_I I~~P I~F '~J I I 1 I IEI'~:'F 1$'Ft1t,! 13n)Ff?E.Z *I .#*.I 3 :I .I . I .I .-I b C DO '1'01-J ItlRtiT TO E: :fiM'ItiE THE F'EGISTEET < t1u I tiP!!T Z TFUC TUPE TO I:\u20ac FRF-ZED I T H I r i l TI-FIT I .IRIll ' .1\"171 IlIlTH HER I THIlib THHT I Z i i l t l '-t'OCI 1111 TH HER lHld t4bT I11 LE%ICOt4 LE: :ICUli A D i l . TO fiE:OF'T-F'HFIE T'I'PE I TDP. E L I \u20ac TS'Pi 'I'EZ 'J'E\" .-Id I2 P-11 ? .-. -AH!?? FE.a $LIFE ,TFIL t4G 1% ! T I TPHtiZ, tTC4 1 F'H 3T > II~Q~;~I$ 3 ZTf3fE Z COlrlPLEMEtfT ? I W I {if:: HEF I :TF'UCTI,IFE: 11if'r'PE DI:LI (:SlJE:J I I~P~P F O I I F T i F F ' E ' S I~~~T l H I l i ) E ' * (OBJ fi-riP11;QMP ~I~T ' I ' F ' E", |
| "uris": null |
| }, |
| "FIGREF58": { |
| "num": null, |
| "type_str": "figure", |
| "text": "N:; lbfi:.;T) 1 } ) 110 YOIJ umg'r rn T X~I~+ T . N K 71-11: ,KG 1:;rlir;S DE I\"')iRSCTi . . I I I I 1 I , I S lil. C;I\\L.CSS I I I 4 I I I 1 1 2 iil-Clil-ESE; E;TttI,E C L f; TF: I E4r ; f 17ECI;LESS LCLl I I.JI :, 1 l i l l l : 1 IJIit-1 t S ( TYIvr: D C L ) (!;LI.T,;J I ( U I * ( T N S iL'PR(;i (V T 14liI:t\\l\\) (OL{-l (i4F1(N IfTSl!ESk)) -THE F{QY R u N N I J~J \\ ]~G TO THE HOUSE I S JOHN 'TtiE BOY RkJNl43 NG T O TliE HOUSE I S J O H N S'1A7E S COrlPLEMENT STRING: JOHN BUILrl S TI?UCTUI7E *' (S(TYF'E UCI-1 (SUGJ OEP(DET TI-IE) (id D o Y ) (EMB (Ul='iU RUN) (TNS F'F'RGI (UFp (I\"RI,r-' T CI (Ef1'-C DE r THE ) ( N I-IUUSE > > > )'> j ) > i P I E D ( VF' ( U BE)< T N S F'RES)(N?I'IP (I!I:'(NE'R J O I -I N ) ) ) ) ) )TI^ ~J u lrJriNr TO E X R M l N E THE REGISTERS ? NO I N P U T STF;UCTURE TO BE PARSED K:REklxING I I I S W C S IS F:ECKLESS D R E k l i l 1.IG I l l S l i E S IS RECKLESS ST'f2T.F S CCIFil='LEMi:I!T S T R I f 4 G t RECliLESS BUILD STRUCTURE t (S(TYF'E D C L i ( 6 U K i J tNP(COM13 CVF(TNS T-'F'RG)(VT EREAh)(OBJ ( N F ' ( N D I S I -I E S l > ) ) ) I ) E L P J E:E > ( T N S PRES) f AT1 J RECIiLESS ) ) 1 110 YOU WANT T O E X A M I N E THE REGISTERS ? NO INPUT STRUCTURE T O BE PARSED I WAS TkJIt4liING THAT Y O U E\\F\\WEKE CONSERVATIVE f U k S THI?.II<ING THAT Y O U WERE CQNSERVATIVE S T A T E SCOMPLEMTi4T STRING! CONSERVATIVEPUIL'It S T R U C T U R E : {SCTYFIE DCI,) (SUFJ (NF'(PR0 1)) 1 (FRET1 CUF'(RUX E:E) ( T N S FF'RG) ( V TI-/INI<j-CFYRPJF' (NF'CCOMI=' ( S (TYF'E ICILL~ ( S U~J ( N F ' ( F R O Y O U ) ) (F'F:ED ( I 3 E E H T N S I='AST)(ATiJ CDNSERVCiT1VE)))I)))))) DO YOU W R r X TUEXAhIIIE THE REGISTEl3S ? NO TNFUT STKUCT-URE TO BE PARSED JOIN w e P E L I~V E D T D R E DELAYEII 401 i V WiiS t : l f L I E Vt~t TO BE DELAYED 5 7 k T t S COMPLEI.SEN T S'Tf I NG t DELAYEX1 BUILP S TRUC'TURE t (8 ( TYF'IE TriPhS ) t S~IIEI J SOHEONE? ( PRECI (:GUX, Blfb> ( TNS P~E S 1 t U *EEL JEVE ) (CIFJES C S Z TYI\"E TT<FI.IS*) (,SUE{ J SC)r?EOl.ll: > I ILZV1.' ( V~?J( A U X Pic) ( T PIIS F'I\"11T 3 ( U DELAY) (C1-E.J ( N F (~t I =~l 7 J Q l { ] g j > > j ) 1 > ) 3 1 DO vDU WRi$'r TO EX#+MI IJE 'THC t7FGTSSTETiSj ? ND rbrrur S T I x u c -r u l x cu EE pAflSED T.Hi?T tie E E-IER IiS 31 1 WAS SI,RI.OWS TI-lhT HE t:fT~liE l 711SR I C 1 I S1.f YhS' SERIOUS-STATE S COMr.'LEHISl4T STRT'NG $ C:ERIOUS B U I L D STI2WCTURE t (S $TYPE lDCld) (WC J (WP( COI.IR I S (TYPE DCL, 1 (SlJp,l ( 1 HE) j I.(F'ftED (V17'fifNS ~A $ ' J ) ( V -f G R E h l i ) (OIIJ (I),F'(FBFiD H~K ) ( N D f S 1 -l ) ) ) ) ) I ) ! ) ) ) W R E I J V E f , F.h:;T) ChLlJ I;Efi,ID,US:) 1 ) DO Y O U WANT i ( 3 .EXtlMINE TtlE REGIGTERs ? ' YES -", |
| "uris": null |
| }, |
| "TABREF1": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td colspan=\"2\">Microfiche 55 : $15.00 ISBN 0-262-11059-8</td></tr><tr><td colspan=\"2\">REVIEWED B Y SIEB NOOTEBOOM</td></tr><tr><td colspan=\"2\">Instituut voor P e r c e p t i e Onderzoek</td></tr><tr><td>Postbus 513 den Dolech 2</td><td>Eindhoven 4502</td></tr></table>", |
| "text": "" |
| }, |
| "TABREF2": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>3 0</td></tr></table>", |
| "text": "THE EVOLUTION OF SPEECH AND LANGUAGE In a number of places in this volume attempts are made to relate r e s u l t s of r e c e n t empirical studies of several kinds to theoretical ideas on the evolution of speech and language in early man. So Peter Pfarler gives an interesting description of communication systems in nonhuman primates and birds. of speech sounds may have directly facilitated the introduction of syntax as a radical innovation in primate communication\". There appear to be two basic assumptions underlying Marler's reasoning. One is that comparative studies of sensory and vocal behavi~r in animals and man maxr lead to interesting theories about specific properties of the human brain underlying man's capacity for speech and language. The other is" |
| }, |
| "TABREF3": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>Dm The Role of S p e e c h in Language sign language Stokoe's</td><td>3G</td></tr></table>", |
| "text": "Ldnguayeinterpretation is that there are feature detectors which can be fatigued by repeated stimulation. By carefully studying which acoustic configurations lead to shifts in particular response distributions, it would be possible to find out what information is extracted by particular feature detectors.Cutting and Eimas argue for the existence of phonetic, speech s p e c i i i c , feature detectors. More recent studies show that categorical perception and selective adaptation are not unique to speech perception (Cutting, Rosner and Foard 1976) . Furthermore, to my knowledge, nobody has yet seriously discussed the. difficulties for a theory of \"wired-in\" feature detectors stemming from perceptual normalization experiments in which it is shown that response distributions in phoneme identification tasks may shift systematically due to the immediate environm e n t of the test segment (e .g . Fourcin 1972) . THE ABSENCE O F PROSODY T h e volume under review is not only remarkable for the many interesting and stimulating papers it contains but also for -what it does not con&ain. In a collection of papers with the title \"The r o l e of speech in language\" one w o~l d have expected to find at least one contribution seriously discussing the relation between speech prosody and linguistic structure. It is ironical that the only paper in which intonational contrast treatment of intonation and its kinesic correlate in sign language seems to make explicit why so many speech researchers do not pay attention to speech prosody. He suggests that intonational contrasts \"are not necessarily linguistic and have more affinity with other systems that signal affect than with phonemic contrasts. There remain then only phonemic contrasts between consonant and consonant, vowel and vowel, and tone and tone (when so used) as the ihdisputably linguistic, basic features of language\". One may fear that this undue overemphasis on phonemic contrast in speech perception research will persist until speech scientists turn away from the study of isolated CV-syllables and start wondering about the perception of normal spontaneous connected speech." |
| }, |
| "TABREF5": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>Term Value Mcasurcments</td></tr></table>", |
| "text": "" |
| }, |
| "TABREF6": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>50 Terms with</td></tr><tr><td>High Discrimination</td></tr><tr><td>Values</td></tr><tr><td>3</td></tr></table>", |
| "text": "" |
| }, |
| "TABREF8": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td/><td>-</td></tr><tr><td>5 0 High Relevance</td><td>50 Low Relevance</td></tr><tr><td>Terms *-B =10.3 --k F -24.6</td><td>Terms -& B =58.9 F =84.0 -A</td></tr><tr><td>3657 t o 420</td><td>1 5 8 4 t~ 4 3 2</td></tr><tr><td>average 38.95%</td><td>average 20.66%</td></tr><tr><td colspan=\"2\">t-test 0.000n2</td></tr></table>", |
| "text": "" |
| }, |
| "TABREF9": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>e c 2 2 . 5</td><td>* F =41.9</td></tr></table>", |
| "text": "Low Relevance Terms" |
| }, |
| "TABREF13": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>PARSER</td></tr></table>", |
| "text": "BACKlooks up the word class of t h e c u r r e n t first word.in the.input string. If t h e word i s not in t h e lexicon an add r o u t i n e is called which permits additions. IfCAT succeeds by matching the current word class with its argument, the word is removed f r a m the input string and pushed ont-o a staclc (SAVEW) . If i t fails an alternate class is tested, provided chat the alternate flag is on. Fail return l e a v e s the surface string unaltered." |
| }, |
| "TABREF14": { |
| "num": null, |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>T O I rd r:r: fi nnbt wn; ; 1 1 1 s 11nEnM n I wnr; t11 . ; rll-iEnM 5 -1.17 T r L; I I I S T R T N G : DIXECiM rw 1 I. I I :+~ITI.JC T tmr: :</td></tr></table>", |
| "text": "I ; ( T?I\"IT DCI, 1 ( ,I;IJL{ J ( N I ' ( I;OMTB (S ( ' T Y[.'E 1 C Z1IK.i J 1 ( ET;VF.' ( VI\" ( V PE ) J r I ( I t I t h j ( E4 MAFI 1 i 1 j 1 ) 1 i ) ( l'*l<L I1 ( VI-'" |
| } |
| } |
| } |
| } |
