Constrained phrase-based translation using weighted finite-state transducers

Abstract

Phrase-based translation models have shown clear advantages over word-based models, and weighted finite-state transducers (WFST's) provide a unified framework for integrating the various components of a speech-to-speech translation system, such as speech recognition and machine translation. This paper combines these two ideas by proposing a constrained phrase-based statistical machine translation system that we implement using WFST's. We evaluate the proposed model on a bidirectional Chinese-English translation task and show improvements over our previous system. © 2005 IEEE.