摘要: LFG-DOP (Bod and Kaplan, 1998, 2003) provides an appealing answer to the question of how probabilistic methods can be incorporated into linguistic theory. However, despite its attractions, standard model suffers from serious problems overgeneration because (a) it is unable define fragments right level generality, (b) has no way capturing effect anything except simple positive constraints. We show extended overcome these problems.
essex.ac.uk
stanford.edu参考文章(20)
Josef van Genabith, Mary Hearne, Andy Way, Ríona Finn, GF-DOP: grammatical feature data-oriented parsing CSLI Publications. ,(2006)
Dan Jurafsky, Probabilistic Modeling in Psycholinguistics: Linguistic Comprehension and Production ,(2006)
Rens Bod, Beyond grammar : an experience-based theory of language ,(1998)
R. Bod, R. Scha, Data-oriented language processing Corpus-Based Methods in Language and Speech. pp. 137- 173 ,(1997) , 10.1007/978-94-017-1183-8_5
Khalil Sima’an, Luciano Buratto, Bacoff parameter estimation for the DOP model european conference on machine learning. ,vol. 2837, pp. 373- 384 ,(2003) , 10.1007/978-3-540-39857-8_34
Ash Asudeh, Ida Toivonen, Stephen Wechsler, Joan Bresnan, Lexical-functional syntax Wiley Blackwell. ,(2015)
R. Bod, R. Kaplan, A DOP model for lexical-functional grammar Data-Oriented Parsing. pp. 211- 233 ,(2003)
Rens Bod, Khalil Sima'an, Remko Scha, Data-Oriented Parsing CSLI Studies in comutational Linguistics. ,(2003)
Kim Honeyford, Mary Dalrymple, Ronald M. Kaplan, John T. Maxwell, Annie Zaenen, Formal issues in lexical-functional grammar Language. ,vol. 74, pp. 231- ,(1998) , 10.2307/417641
Mark Johnson, Squibs and discussions: the DOP Estimation method is biased and inconsistent Computational Linguistics. ,vol. 28, pp. 71- 76 ,(2002) , 10.1162/089120102317341783