#### DMCA

## Generation and Synchronous Tree-Adjoining Grammars (1990)

### Cached

### Download Links

- [acl.ldc.upenn.edu]
- [people.deas.harvard.edu]
- [www.eecs.harvard.edu]
- [ucrel.lancs.ac.uk]
- [www.aclweb.org]
- [aclweb.org]
- [wing.comp.nus.edu.sg]
- [www.aclweb.org]
- [www.eecs.harvard.edu]
- [wing.comp.nus.edu.sg]
- [acl.ldc.upenn.edu]
- [www.aclweb.org]
- [www.aclweb.org]
- [aclweb.org]
- [aclweb.org]
- [ucrel.lancs.ac.uk]
- [www.aclweb.org]
- [www.seas.upenn.edu]
- [www.seas.upenn.edu]
- [www.cis.upenn.edu]
- [semarch.linguistics.fas.nyu.edu]
- [ucrel.lancs.ac.uk]
- [newdesign.aclweb.org]
- [www.aclweb.org]
- [aclweb.org]
- [acl.ldc.upenn.edu]
- [repository.upenn.edu]
- [www.deas.harvard.edu]
- [www.eecs.harvard.edu]
- [repository.upenn.edu]
- [www.cis.upenn.edu]
- [www.seas.upenn.edu]
- [www.seas.upenn.edu]

Citations: | 774 - 43 self |

### Citations

10912 |
A mathematical theory of communication
- Shannon
- 1948
(Show Context)
Citation Context ...hastic techniques applied to syntax modeling have recently regained popularity, current language models suffer from obvious inherent inadequacies. Early proposals such as Markov Models, N-gram models =-=[1, 2, 3]-=- and Hidden Markov Models were very quickly shown to be linguistically not appropriate for natural language (e.g. [4]) since they are unable to capture long distance dependencies or to describe hierar... |

798 | An efficient context-free parsing algorithm.
- Earley
- 1970
(Show Context)
Citation Context ...non-terminal symbol. The algorithm collects items into a chart. 21 An item s is de ned as an 8-tuple, s =[ ; dot; pos; i; j; k; l; sat?] where: 21 We could have grouped the items into item sets as in =-=[Earley1968]-=- but we chose not to, allowing us to de ne an agenda driven parser.40 Joshi and Schabes Start End A A B C D B C D E F G H I E F G H I Fig. 10.5. Left, left to right tree traversal; right, equivalent ... |

609 | Lexical-functional grammar: A formal system for grammatical representation.
- Kaplan, Bresnan
- 1982
(Show Context)
Citation Context ...stree on the left in Figure 1):sinterior nodes are labeled by non-terminal symbols;slSome of the linguistic formalisms illustrating the increased use of lexical information are, lexical rulessin LFG (=-=Kaplan and Bresnan, 1983-=-), GPSG (Gazdar et al., 1985), HPSG (Pollard and Sag, 1987), Combi-snatory Categorial Grammars (Steedman, 1987), Karttunen's version of Categorial Grammar (Karttunen,s1986), some versions of GB theory... |

573 | The mathematics of sentence structure’. - Lambek - 1958 |

551 |
Generalized phrase structure grammar.
- Gazdar, Klein, et al.
- 1985
(Show Context)
Citation Context ...nterior nodes are labeled by non-terminal symbols;slSome of the linguistic formalisms illustrating the increased use of lexical information are, lexical rulessin LFG (Kaplan and Bresnan, 1983), GPSG (=-=Gazdar et al., 1985-=-), HPSG (Pollard and Sag, 1987), Combi-snatory Categorial Grammars (Steedman, 1987), Karttunen's version of Categorial Grammar (Karttunen,s1986), some versions of GB theory (Chornsky, 1981)), and Lexi... |

486 |
Prediction and entropy of printed English
- Shannon
- 1951
(Show Context)
Citation Context ...hastic techniques applied to syntax modeling have recently regained popularity, current language models suffer from obvious inherent inadequacies. Early proposals such as Markov Models, N-gram models =-=[1, 2, 3]-=- and Hidden Markov Models were very quickly shown to be linguistically not appropriate for natural language (e.g. [4]) since they are unable to capture long distance dependencies or to describe hierar... |

429 |
The estimation of stochastic context-free grammars using the insideoutside algorithm.
- Lari, Young
- 1990
(Show Context)
Citation Context ...ncies or to describe hierarchically the syntax of natural languages. Stochastic context-free grammar [5] is a hierarchical model more appropriate for natural languages, however none of such proposals =-=[6, 7]-=- perform as well as the simpler Markov Models because of the difficulty of capturing lexical information. The parameters of a stochastic context-free grammar do not correspond directly to a distributi... |

387 | Tree Adjunct Grammars. - Joshi, Levy, et al. - 1975 |

306 | Inside-outside reestimation from partially bracketed corpora.
- Pereira, Schabes
- 1992
(Show Context)
Citation Context ... for estimating the parameters of a SLTAG given a training corpus of raw text. Similarly to its context-free counterpart, the reestimation algorithm can be extended to handle partially parsed corpora =-=[20]-=-. The worst case complexity of the algorithm with respect to the length of the input string (O(N6)) makes it impractical with a large corpus on a single processor computer for grammars requiring the w... |

301 |
Trainable grammars for speech recognition.
- Baker
- 1979
(Show Context)
Citation Context ...algorithm derives a new estimate such that the probability that the corpus were generated by the grarnlnar is increased. By analogy to the inside-outside algorithm for stochastic contextfree grammars =-=[19, 7]-=-, we believe that the following quantity decreases after each iteration: 1° log2(P(w)) He(W) = to¢w (15) toEW In order to derive a new estimate, the algorithm needs to compute for all sentences in W t... |

166 |
Mathematical and computational aspects of lexicalized grammars.
- Schabes
- 1990
(Show Context)
Citation Context ...tive and are therefore more appropriate for statistical analysis of language. In fact, LTAGs are the simplest hierarchical formalism which can serve as the basis for lexicalizing context-free grammar =-=[10, 11]-=-. LTAG is a tree-rewriting system that combines trees of large domain with adjoining and substitulion. The trees found in a TAG take advantage of the available extended domain of locality by localizin... |

158 |
Characterizing Mildly Context-Sensitive Grammar Formalisms.
- Weir
- 1988
(Show Context)
Citation Context ... TAGs describing their linguistic, computational and formal properties. Some of these are: Joshi [Joshi1987], Joshi, Vijay-Shanker and Weir [Joshi et al.1991], Vijay-Shanker [Vijay-Shanker1987], Weir =-=[Weir1988]-=-, Schabes [Schabes1990, Schabes1991], Schabes and Joshi [Schabes and Joshi1988, Schabes and Joshi1989], Kroch [Kroch1987], Kroch and Joshi [Kroch and Joshi1985], Abeille, Bishop, Cote and Schabes [Abe... |

149 | The convergence of mildly context-sensitive grammar formalisms
- Joshi, Shanker, et al.
- 1991
(Show Context)
Citation Context ... to other grammatical formalisms such as combinatory categorial grammars and modified head grammars since they have been proven to be equivalent to tree-adjoining grammars and linear indexed grammars =-=[21]-=-. In collaboration with Aravind Joshi, Fernando Pereira and Stuart Shieber, we are currently investigating additional algorithms and applications for SLTAG, methods for lexical clustering and automati... |

126 |
How much context-sensitivity is necessary for characterizing structural descriptions.
- Joshi
- 1985
(Show Context)
Citation Context ...Tree- Adjoining GrammarssTAGs were introduced by Joshi, Levy and Takahashi (1975) and Joshi (1985). For moresdetails on the original definition of TAGs, we refer the reader to (Joshi, 1987; Kroch ands=-=Joshi, 1985-=-). It is known that tree-adjoining languages (TALs) generate some strictlyscontext-sensitive languages and fall in the class of the so-called 'mildly context-sensitiveslanguages' (Joshi et al., forthc... |

126 | Parsing strategies with “lexicalized” grammars: Application to tree adjoining grammars.
- Schabes, Abeille, et al.
- 1988
(Show Context)
Citation Context ...G given a training corpus of text. This algorithm can either be used for refining the parame1 We assume familiarity throughout the paper with TAGs and its lexicalized variant. See, for instance, [S], =-=[9]-=-, [10] or [111. 140ters of a SLTAG or for inferring a tree-adjoining grammar from a training corpus. Due to the lack of space, in this paper the algorithms are described succinctly without proofs of ... |

125 | An alternative conception of tree-adjoining derivation - Schabes, Shieber - 1992 |

124 |
Basic methods of probabilistic context-free grammars, in:
- Jelinek, Lafferty, et al.
- 1992
(Show Context)
Citation Context ...ncies or to describe hierarchically the syntax of natural languages. Stochastic context-free grammar [5] is a hierarchical model more appropriate for natural languages, however none of such proposals =-=[6, 7]-=- perform as well as the simpler Markov Models because of the difficulty of capturing lexical information. The parameters of a stochastic context-free grammar do not correspond directly to a distributi... |

123 | Disambiguation of super parts of speech (or supertags): Almost parsing - Joshi, Srinivas - 1994 |

121 |
A Study of Tree Adjoining Grammars
- Vijay-Shanker
- 1987
(Show Context)
Citation Context ...ree languages also hold for tree-adjoining languages. In fact, TALs are a full abstract family of languages (full AFLs). avariant of the push-down automaton called embedded push-down automaton (EPDA) =-=[Vijay-Shanker1987]-=- characterizes exactly the set of tree-adjoining languages, just as push-down automaton characterizes CFLs. there is a pumping lemma for tree-adjoining languages. tree-adjoining languages can be parse... |

113 | The linguistic relevance of tree adjoining grammar,
- Kroch, Joshi
- 1985
(Show Context)
Citation Context ...at-sical formalism with attractive properties for the strongscharacterization of the syntax of natural languages, thatsis, characterization of the analysis trees of the expres-ssions in the language (=-=Kroch and Joshi, 1985-=-; Kroch,s1989).l Among these properties are thatsThe domain of locality in TAGs is larger thansfor formalisms that augment context-free grammarss(such as lexical-functional, or generalized or head-sdr... |

107 | D-tree grammars - Rambow, Vijay-Shanker, et al. - 1995 |

103 |
Treeadjoining grammars and lexicalized grammars.
- Joshi, Schabes
- 1991
(Show Context)
Citation Context ...tive and are therefore more appropriate for statistical analysis of language. In fact, LTAGs are the simplest hierarchical formalism which can serve as the basis for lexicalizing context-free grammar =-=[10, 11]-=-. LTAG is a tree-rewriting system that combines trees of large domain with adjoining and substitulion. The trees found in a TAG take advantage of the available extended domain of locality by localizin... |

101 | An efficient recognition and syntax algorithm for context-free languages. Scientific Report AFCRL-65-758, Air Force Cambridge Research Lab., - Kasami - 1965 |

100 |
A uniform architecture for parsing and generation
- Shieber
- 1988
(Show Context)
Citation Context ...B et al., 1990).sIn previous work, one of us noted that generation ac-scording to an augmented context-free grammar can besmade more efficient by requiring the grammar to be se-smantically monotonic (=-=Shieber, 1988-=-); the derived se-smantics for an expression must include, in an appropri-sate sense, the semantic material of all its subconstituents.sIt is interesting to note that synchronous TAGs are in-sherently... |

97 |
Applicability of Indexed Grammars to Natural Languages
- Gazdar
- 1988
(Show Context)
Citation Context ...), as a notation for SLTAGs. A linear indexed grammar is constructed following the method given in [13]. However, in addition, each rule is associated with a probability. Linear Indexed grammar (LIG) =-=[14, 15]-=- is a rewriting system in which the non-terminal symbols are augmented with a stack. In addition to rewriting non-terminals, the rules of the grammar can have the effect of pushing or popping symbols ... |

94 | Formal and Computational Aspects of Natural Language Syntax
- Rambow
- 1994
(Show Context)
Citation Context ...habes [Abeille et al.1990], Abeille [Abeille1988], Schabes and Waters [Schabes and Waters1995], Rambow, Vijay-Shanker and Weir [Rambow et al.1995], Joshi and Srinivas [Joshi and Srinivas1994], Rambow =-=[Rambow1994]-=-, Vijay-Shanker [Vijay-Shanker1992], Shieber and Schabes [Shieber and Schabes1990]. A reader interested in TAGs will nd these papers very useful. Additional useful references will be found in the Spec... |

92 | An algorithm for generating quantifier scopings - Hobbs, Schieber - 1987 |

87 |
Combinatory grammars and parasitic gaps. Natural Language and Linguistic Theory
- STEEDMAN
- 1987
(Show Context)
Citation Context ...malisms illustrating the increased use of lexical information are, lexical rules in LFG [Kaplan and Bresnan1983], GPSG [Gazdar et al.1985], HPSG [Pollard and Sag1987], Combinatory Categorial Grammars =-=[Steedman1987]-=-, Karttunen's version of Categorial Grammar [Karttunen1986], some versions of GB theory [Chomsky1981], and LexiconGrammars [Gross1984].Tree-Adjoining Grammars 3 (i) is a nite set of terminal symbols;... |

82 | Using description of trees in a tree adjoining grammar
- Vijay-Shanker
- 1992
(Show Context)
Citation Context ...Abeille [Abeille1988], Schabes and Waters [Schabes and Waters1995], Rambow, Vijay-Shanker and Weir [Rambow et al.1995], Joshi and Srinivas [Joshi and Srinivas1994], Rambow [Rambow1994], Vijay-Shanker =-=[Vijay-Shanker1992]-=-, Shieber and Schabes [Shieber and Schabes1990]. A reader interested in TAGs will nd these papers very useful. Additional useful references will be found in the Special Issue of Computational Intellig... |

81 |
A.K.: Some computational properties of tree adjoining grammars
- Vijay-Shanker, Joshi
- 1985
(Show Context)
Citation Context ...ABILITY OF A SENTENCE We now define an bottom-up algorithm for SLTAG which computes the probability of an input string. The algorithm is an extension of the CKY-type parser for tree-adjoining grammar =-=[18]-=-. The extended algorithm parses all spans of the input string and also computes their probability in a bottom-up fashion. Since the string on the frontier of an auxiliary is broken up into two substri... |

73 |
Radical Lexicalism.
- Karttunen
- 1989
(Show Context)
Citation Context ...ion are, lexical rules in LFG [Kaplan and Bresnan1983], GPSG [Gazdar et al.1985], HPSG [Pollard and Sag1987], Combinatory Categorial Grammars [Steedman1987], Karttunen's version of Categorial Grammar =-=[Karttunen1986]-=-, some versions of GB theory [Chomsky1981], and LexiconGrammars [Gross1984].Tree-Adjoining Grammars 3 (i) is a nite set of terminal symbols; (ii) NT is a nite set of non-terminal symbols 2 : \ NT = ;... |

65 | Feature Structures Based Tree Adjoining Grammars - Vijay-Shanker, Joshi - 1988 |

64 | Recognition and Parsing of Context-Free Languages - Younger - 1967 |

64 |
The Mathematics of Sentence Structure. The American Mathematical Monthly
- Lambek
- 1958
(Show Context)
Citation Context ... components of their arguments. Wescould also impose that the operations yield languages of constant growth (Joshi [1985]).sThe operations that we will use have these properties.sCategorial Grammars (=-=Lambek, 1958-=-; Steedman, 1987) are lexicalized according tosour definition since each basic category has a lexical item associated with it.sAs in Categorial Grammars, we say that the category of a word is the enti... |

55 |
Information-based Syntax and Semantics: Vol 1: Fundamentals. Center for the Study of Language and Information
- Pollard, Sag
- 1988
(Show Context)
Citation Context ... non-terminal symbols;slSome of the linguistic formalisms illustrating the increased use of lexical information are, lexical rulessin LFG (Kaplan and Bresnan, 1983), GPSG (Gazdar et al., 1985), HPSG (=-=Pollard and Sag, 1987-=-), Combi-snatory Categorial Grammars (Steedman, 1987), Karttunen's version of Categorial Grammar (Karttunen,s1986), some versions of GB theory (Chornsky, 1981)), and Lexicon-Grammars (Gross, 1984).s'W... |

51 | The relevance of tree adjoining grammar to generation - Joshi - 1987 |

51 |
Probabilistic representation of formal languages.
- Booth
- 1969
(Show Context)
Citation Context ... appropriate for natural language (e.g. [4]) since they are unable to capture long distance dependencies or to describe hierarchically the syntax of natural languages. Stochastic context-free grammar =-=[5]-=- is a hierarchical model more appropriate for natural languages, however none of such proposals [6, 7] perform as well as the simpler Markov Models because of the difficulty of capturing lexical infor... |

42 | An Earley-type parsing algorithm for Tree Adjoining Grammars, 26 th Annual Meeting of the Association for Computational Linguistics - Schabes, Joshi - 1988 |

41 | An algorithm for generation in unification categorial grammar - Calder, Reape, et al. - 1989 |

36 |
Lexicon-grammar and the syntactic analysis of French
- Gross
- 1984
(Show Context)
Citation Context ...], HPSG [Pollard and Sag1987], Combinatory Categorial Grammars [Steedman1987], Karttunen's version of Categorial Grammar [Karttunen1986], some versions of GB theory [Chomsky1981], and LexiconGrammars =-=[Gross1984]-=-.Tree-Adjoining Grammars 3 (i) is a nite set of terminal symbols; (ii) NT is a nite set of non-terminal symbols 2 : \ NT = ;; (iii) S is a distinguished non-terminal symbol: S 2 NT; (iv) I is a nite ... |

33 | Parsing with Lexicalized Tree Adjoining Grammar - Schabes, AK - 1990 |

33 | Reversible logic grammars for machine translation - Dymetman, lsabelle - 1988 |

29 | A lexicalized tree adjoining grammar for English - Abeille, Bishop, et al. - 1990 |

28 |
Unbounded Dependencies and Subjacency in a Tree Adjoining Grammar
- Kroch
- 1987
(Show Context)
Citation Context ...ay-Shanker and Weir [Joshi et al.1991], Vijay-Shanker [Vijay-Shanker1987], Weir [Weir1988], Schabes [Schabes1990, Schabes1991], Schabes and Joshi [Schabes and Joshi1988, Schabes and Joshi1989], Kroch =-=[Kroch1987]-=-, Kroch and Joshi [Kroch and Joshi1985], Abeille, Bishop, Cote and Schabes [Abeille et al.1990], Abeille [Abeille1988], Schabes and Waters [Schabes and Waters1995], Rambow, Vijay-Shanker and Weir [Ram... |

24 | Tags as a grammatical formalism for generation. - McDonald, Pustejovsky - 1985 |

24 |
A Lexicalized Tree Adjoining Grammar for English
- XTAG-Group
- 1998
(Show Context)
Citation Context ...voted to TreeAdjoining Grammars. A wide coverage lexicalized TAG grammar for English (about 300,000 in ected items and about 570 trees in 38 families) and a parser (XTAG System) has been described in =-=[XTAG-Group1995]-=-, which includes evaluation of XTAG on corpora such as the Wall Street Journal, IBM Computer Manuals and ATIS Corpus.Tree-Adjoining Grammars 23 is such a class. This class however cannot capture cont... |

23 | Deterministic left to right parsing of tree adjoining languages - Schabes, Vijay-Shanker - 1990 |

22 | Tree-adjoining grammar parsing and Boolean matrix multiplication
- Satta
- 1994
(Show Context)
Citation Context ...local or global ambiguity. Therefore, when an item is added, if the item is already in the chart, the new items that caused the same item to exist are added to the set of causes. 24 23 Recently Satta =-=[Satta1994]-=- was able to transfer the complexity bound of TAG parsing to the one of matrix multiplication. As a consequence, it is shown that if one were to improve the bound of O(n 6 )-time for the TAG parsing p... |

21 | Bidirectional parsing of lexicalized tree adjoining grammar - Lavelli, Satta - 1991 |

20 | The valid prefix property and left to right parsing of tree-adjoining grammar - Schabes - 1991 |

16 |
Indexed grammars, an extension of context free grammars," Princeton University Doctoral Thesis
- Aho
- 1967
(Show Context)
Citation Context ...), as a notation for SLTAGs. A linear indexed grammar is constructed following the method given in [13]. However, in addition, each rule is associated with a probability. Linear Indexed grammar (LIG) =-=[14, 15]-=- is a rewriting system in which the non-terminal symbols are augmented with a stack. In addition to rewriting non-terminals, the rules of the grammar can have the effect of pushing or popping symbols ... |

13 | Asymmetries in Long Distance Extraction in a TAG Grammar - Kroch - 1989 |

13 | The Systematic Construction of Earley Parsers: Application to the Production of O(n6) Earley Parsers for Tree Adjoining Grammars - Lang - 1988 |

13 |
Characterizing derivations trees of context free grammars through a generalization of finite automata theory
- Thatcher
- 1971
(Show Context)
Citation Context ...can be parsed in polynomial time, in the worst case in O(n6 ) time. Some well know properties of the tree sets of tree-adjoining grammars follow: the tree sets of recognizable sets (regular tree sets)=-=[Thatcher1971]-=- are strictly included in the tree sets of tree-adjoining grammars, T (TAG); recognizable sets T(TAG) 5 Wesay that an initial tree is completed if there is no substitution nodes on the frontier of it.... |

12 | Parsing French with tree-adjoining grammar: Some linguistic accounts
- Abeillé
- 1988
(Show Context)
Citation Context ...habes1991], Schabes and Joshi [Schabes and Joshi1988, Schabes and Joshi1989], Kroch [Kroch1987], Kroch and Joshi [Kroch and Joshi1985], Abeille, Bishop, Cote and Schabes [Abeille et al.1990], Abeille =-=[Abeille1988]-=-, Schabes and Waters [Schabes and Waters1995], Rambow, Vijay-Shanker and Weir [Rambow et al.1995], Joshi and Srinivas [Joshi and Srinivas1994], Rambow [Rambow1994], Vijay-Shanker [Vijay-Shanker1992], ... |

11 |
Tree pushdown automata
- Schimpf, Gallier
- 1985
(Show Context)
Citation Context ..., /P(T(G)), is a context-sfree language;sP(T(G))sis a CFLsr the tree sets of TAG are equivalent to the tree sets of linear indexed languages.sHence, linear versions of Schimpf-Gallier tree automaton (=-=Schimpf and Gallier, 1985-=-)sare equivalent to 'T(TAG);sr for every TAG, G, the tree set of G, T(G), is recognizable in polynomial time, insthe worst case in O(n3)-time, where n is the number of nodes in a tree t E T(G).sWe now... |

9 | The relevance of lexicalization to parsing - Schabes, Joshi - 1989 |

9 | Automated Inversion of a Unification Parser into a Unification Generator - Strzalkowski - 1989 |

8 | Parsing Idioms in Tree Adjoining Grammars - AbeillC, Schabes - 1989 |

7 | Tree insertion grammar: A cubic time, parsable formalism that lexicalizes context-free grammars without changing the trees produced - Schabes, Waters - 1995 |

7 |
Parsing Constrained Grammar Formalisms
- Vijay-Shanker, Weir
- 1993
(Show Context)
Citation Context ...n and elegance, we use a stochastic linear rewriting system, stochastic linear indexed grammars (SLIG), as a notation for SLTAGs. A linear indexed grammar is constructed following the method given in =-=[13]-=-. However, in addition, each rule is associated with a probability. Linear Indexed grammar (LIG) [14, 15] is a rewriting system in which the non-terminal symbols are augmented with a stack. In additio... |

7 |
An inside-outside algorithm for estimating the parameters of a hidden stochastic context-free grammar based on Earley's algorithm. Unpublished mss
- Schabes
- 1991
(Show Context)
Citation Context ...-adjoining grammar. 4The algorithms explained in this paper can be generalized to lexicalized tree-adjoining grammars that need not be in Chomsky Normal Form using techniques similar the one found in =-=[17]-=-. 141the spine (i.e. subsumes the foot node), include: bE-0] P -I (2) Since (2) encodes an immediate domination link defined by t]he tree-adjoining grammar, its associated probability is one. • Simil... |

6 | A lexicalized tree adjoining grammar for english - Abeill6, Bishop, et al. - 1990 |

6 |
Tree adjunct grammars. Journal of computer and system sciences
- Joshi, Levy, et al.
- 1975
(Show Context)
Citation Context ...orthand for OA(A).sIf there are no substitution nodes in the elementary trees and if there are no constraintsson adjoining, then we have the 'pure' (old) Tree Adjoining Grammar (TAG) as describedsin (=-=Joshi et al., 1975-=-).sThe operation of substitution and the constraints on adjoining are both needed forslinguistic reasons. Constraints on adjoining are also needed for formal reasons in order tosobtain some closure pr... |

5 | Long-distance dependencies as a case of functional uncertainty - KAPLAN, ZAENEN - 1988 |

5 |
Secret and urgent: The story of codes and ciphers. Blue Ribbon Books
- Pratt
- 1939
(Show Context)
Citation Context ...hastic techniques applied to syntax modeling have recently regained popularity, current language models suffer from obvious inherent inadequacies. Early proposals such as Markov Models, N-gram models =-=[1, 2, 3]-=- and Hidden Markov Models were very quickly shown to be linguistically not appropriate for natural language (e.g. [4]) since they are unable to capture long distance dependencies or to describe hierar... |

5 |
Lexicalized tree-adjoining grammar for distributional analysis
- Resnik
- 1991
(Show Context)
Citation Context ...s of a predicative element, the distribution of the adverbs licensed by a specific verb, or the adjectives licensed by a specific noun. This kind of statistical analysis as independently suggested in =-=[12]-=- can be made with LTAGs because of their extended domain of locality but also because of their lexicalized property. In this paper, this intuition is made formally precise by defining the notion of a ... |

3 | Unbounded dependencies in tags and lfg: functional uncertainty is a corolary in tags - Joshi, Vijay-Shanker - 1989 |

3 | Probabilistic tree-adjoining grammars as framework for statistical natural language processing - Resnick - 1992 |

2 |
Syntactic Structures, chapter
- Chomsky
- 1964
(Show Context)
Citation Context ... inherent inadequacies. Early proposals such as Markov Models, N-gram models [1, 2, 3] and Hidden Markov Models were very quickly shown to be linguistically not appropriate for natural language (e.g. =-=[4]-=-) since they are unable to capture long distance dependencies or to describe hierarchically the syntax of natural languages. Stochastic context-free grammar [5] is a hierarchical model more appropriat... |

1 | Parsing strategies with 'lexicalized' grammars: Application to tree adjoining grammars - Shieber, Noord, et al. - 1988 |

1 | Left to right parsing of tree-adjoining grammars - Schabes - 1991 |

1 |
l) = 1 if i = i + 1 and if a = wi +a, 0 otherwise
- If
(Show Context)
Citation Context |

1 |
Ty,i,j,k,l) = 1 ifi =j and if k
- IfT, I
(Show Context)
Citation Context |

1 |
if 7 is a node which does not subsume the foot node of the elementary tree it belongs to
- finally
(Show Context)
Citation Context ...ncies or to describe hierarchically the syntax of natural languages. Stochastic context-free grammar [5] is a hierarchical model more appropriate for natural languages, however none of such proposals =-=[6, 7]-=- perform as well as the simpler Markov Models because of the difficulty of capturing lexical information. The parameters of a stochastic context-free grammar do not correspond directly to a distributi... |

1 | Unbounded dependencies in tags and Kg: functional uncertainty is a corolary in tags - Joshi, Vijay-Shanker - 1989 |

1 | Parsing french with tree adjoining grammar: some linguistic accounts - Abeillh - 1988 |