This paper proposes a method of domain sensitive example based machine translation system between Chinese and English. First, the example base is automatically extracted according to the relative position of word alignments, which is a language independent approach. Then, through combining the text-classification technique, the EBMT system manages to select the most appropriate example base for the input text. It is indicated by the experiments that the method is sensitive to the translation domains and capable of improving the translation performance in a multi-domain environment.
展开▼
