Research on Feature-Based Word Automatic Translation Technology in Japanese-Chinese Translation System

For the limitation of terms translation in Japanese-Chinese machine translation system of dictionaries, this paper studies the technology of Japanese-Chinese terms automatic translation. IPC classification information is added to the translation model as a domain information feature of terms, and as one of the features, it participates in the calculation of translation cost. Then, the relative conditional entropy information with word collocation information is taken as a feature and combined with other features to participate in the calculation of translation model, to enhance the distinction between the pros and cons of translation candidates. Finally, an automatic Japanese-Chinese terms translation system based on multi-features is realized. The experimental results show that the terms translation method based on multi-feature fusion is effective. The bilingual terms dictionary can be established with the help of this system, which improves the accuracy of Japanese-Chinese scientific and technological literature translation.