Non-Local Modeling with a Mixture of PCFGs

TitleNon-Local Modeling with a Mixture of PCFGs
Publication TypeConference Paper
Year of Publication2006
AuthorsPetrov, S., Barrett L., & Klein D.
Published inProceedings of the 10th International Conference on Conference of Computational Natural Language Learning (CoNLL-X '06)
Page(s)14-20
Other Numbers3263
Abstract

While most work on parsing with PCFGs has focused on local correlations between tree configurations, we attempt to model non-local correlations using a finite mixture of PCFGs. A mixture grammar fit with the EM algorithm shows improvement over a single PCFG, both in parsing accuracy and in test data likelihood. We argue that this improvement comes from the learning of specialized grammars that capture non-local correlations.

URLhttp://www.icsi.berkeley.edu/pubs/ai/nonlocalmodeling06.pdf
Bibliographic Notes

Proceedings of the 10th International Conference on Conference of Computational Natural Language Learning (CoNLL-X '06), pp. 14-20, New York, New York

Abbreviated Authors

S. Petrov, L. Barrett, and D. Klein

ICSI Research Group

AI

ICSI Publication Type

Article in conference proceedings