Unsupervised Bilingual POS Tagging with Markov Random Fields

Desai Chen, Shay Cohen, Noah A. Smith, Chris Dyer

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we give a treatment to the problem of bilingual part-of-speech induction with parallel data. We demonstrate that naive optimization
of log-likelihood with joint MRFs suffers from a severe problem of local maxima,
and suggest an alternative – using contrastive estimation for estimation of the parameters. Our experiments show that estimating the parameters this way, using overlapping features with joint MRFs performs better than previous work on the 1984 dataset.
Original languageEnglish
Title of host publicationEMNLP '11 Proceedings of the First Workshop on Unsupervised Learning in NLP
PublisherAssociation for Computational Linguistics
Pages64-71
Number of pages8
ISBN (Print)978-1-937284-13-8
Publication statusPublished - 2011

Fingerprint

Dive into the research topics of 'Unsupervised Bilingual POS Tagging with Markov Random Fields'. Together they form a unique fingerprint.

Cite this