Results 1 
6 of
6
Parsing with the Shortest Derivation
 Proceedings COLING2000
, 2000
"... tens @ scs.lecd s.ac.uk Common wisdom has it that tile bias of stochastic grammars in favor of shorter deriwttions of a sentence is hamfful and should be redressed. We show that the common wisdom is wrong for stochastic grammars that use elementary trees instead o1 ' conlextl'ree rules, such as Sto ..."
Abstract

Cited by 36 (14 self)
 Add to MetaCart
tens @ scs.lecd s.ac.uk Common wisdom has it that tile bias of stochastic grammars in favor of shorter deriwttions of a sentence is hamfful and should be redressed. We show that the common wisdom is wrong for stochastic grammars that use elementary trees instead o1 ' conlextl'ree rules, such as Stochastic TreeSubstitution Grammars used by DataOriented Parsing models. For such grammars a nonprobabilistic metric based on tile shortest derivation outperforms a probabilistic metric on the ATIS and OVIS corpora, while it obtains competitive results on the Wall Street Journal (WSJ) corpus. This paper also contains the first publislmd experiments with DOP on the WSJ. 1.
Automatic FStructure Annotation Of Treebank Trees
 THE FIFTH INTERNATIONAL CONFERENCE ON LEXICALFUNCTIONAL GRAMMAR, THE UNIVERSITY OF CALIFORNIA AT BERKELEY, 19 JULY  20 JULY 2000, CSLI
, 2000
"... We describe a method that automatically induces LFG fstructures from treebank tree representations, given a set of fstructure annotation principles that define partial, modular c to fstructure correspondences in a linguistically informed, principlebased way. ..."
Abstract

Cited by 27 (6 self)
 Add to MetaCart
We describe a method that automatically induces LFG fstructures from treebank tree representations, given a set of fstructure annotation principles that define partial, modular c to fstructure correspondences in a linguistically informed, principlebased way.
DataOriented Models of Parsing and Translation
, 2005
"... A dissertation submitted in fulfilment of the requirements for the award of ..."
Abstract

Cited by 12 (2 self)
 Add to MetaCart
A dissertation submitted in fulfilment of the requirements for the award of
An Improved Parser for DataOriented LexicalFunctional Analysis
 In Proceedings of the 38th Conference of the Association for Computational Linguistics
"... We present an LFGDOP parser which uses fragments from LFGannotated sentences to parse new sentences. Experiments with the Verbmobil and Homecentre corpora show that (1) Viterbi n best search performs about 100 times faster than Monte Carlo search while both achieve the same accuracy; (2) the DOP h ..."
Abstract

Cited by 8 (4 self)
 Add to MetaCart
We present an LFGDOP parser which uses fragments from LFGannotated sentences to parse new sentences. Experiments with the Verbmobil and Homecentre corpora show that (1) Viterbi n best search performs about 100 times faster than Monte Carlo search while both achieve the same accuracy; (2) the DOP hypothesis which states that parse accuracy increases with increasing fragment size is confirmed for LFGDOP; (3) LFGDOP's relative frequency estimator performs worse than a discounted frequency estimator; and (4) LFGDOP significantly outperforms TreeDOP if evaluated on tree structures only. 1
LFGDOT: a Probabilistic, ConstraintBased Model for Machine Translation
, 2000
"... We develop novel models for Machine Translation (MT) based on DataOriented Parsing (DOP: Bod, 1995; 1998) allied to the syntactic representations of Lexical Functional Grammar (LFG: Kaplan & Bresnan, 1982). ..."
Abstract
 Add to MetaCart
We develop novel models for Machine Translation (MT) based on DataOriented Parsing (DOP: Bod, 1995; 1998) allied to the syntactic representations of Lexical Functional Grammar (LFG: Kaplan & Bresnan, 1982).
Solving Headswitching Translation Cases in LFGDOT
, 2001
"... these problematic constructions using approaches based on linear logic (Van Genabith et al., 1998) and restriction (Kaplan & Wedekind, 1993), we point out further problems which are introduced. We then show how LFGDOP (Bod & Kaplan, 1998) can be extended to serve as a novel hybrid model for MT, LFG ..."
Abstract
 Add to MetaCart
these problematic constructions using approaches based on linear logic (Van Genabith et al., 1998) and restriction (Kaplan & Wedekind, 1993), we point out further problems which are introduced. We then show how LFGDOP (Bod & Kaplan, 1998) can be extended to serve as a novel hybrid model for MT, LFGDOT (Way, 1999, 2001), which promises to improve upon the DOT model of translation (Poutsma 1998, 2000) as well as LFGMT. LFGDOT improves the robustness of LFGMT through the use of the LFGDOP Discard operator, which produces generalized fragments by discarding certain fstructure features. LFGDOT can, therefore, deal with illformed or previously unseen input where LFGMT cannot. Finally, we demonstrate that LFGDOT can cope with such translational phenomena which prove problematic for other LFGbased models of translation. 1 Headswitching in LFGMT Kaplan et al. (1989) illustrate their LFGMT proposal with the wellknown headswitching case venir de X just Xed, as in (1): has (1) The baby just fell Le bébé vient de tomber. They propose to deal with such problems in two ways. The first of these is as in (2): (2) just: ( PRED) = ‘just ARG ’, ( PRED) = venir, ( XCOMP) = ARG) That is, the XCOMP function of venir (in (1), de tomber) corresponds to the ARG function of just (in (1), the baby fell), as shown by the respective source and target fstructures in (3) and (4): PRED ‘just [fall] ’ (3)