The CASIA phrase-based machine translation system

In this paper we propose a phrase-based translation system. In the system, we use phrase translation model instead of word-based model. An improved method to compute phrase translation probability is studied. A phrase-based decoder we developed employs a beam search algorithm, in which some target language words that have both high frequency of appearance and also fertility zero are introduced to make the result more reasonable. We improve the previously proposed tracing back algorithm to get the best path. Some experiments concerned are presented.