Paper: Nonparametric Bayesian Machine Transliteration with Synchronous Adaptor Grammars

ACL ID P11-2094
Title Nonparametric Bayesian Machine Transliteration with Synchronous Adaptor Grammars
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2011
Authors

Machine transliteration is defined as auto- matic phonetic translation of names across languages. In this paper, we propose syn- chronous adaptor grammar, a novel nonpara- metric Bayesian learning approach, for ma- chine transliteration. This model provides a general framework without heuristic or re- striction to automatically learn syllable equiv- alents between languages. The proposed model outperforms the state-of-the-art EM- based model in the English to Chinese translit- eration task.