Conclusion

In this paper, we discribed pattern-based English-Japanese machine translation with statistical method. And we showed how to automatically create English-Japanese statistical pattern-based machine translation. Ordinarily, translation patterns are made by hand in pattern-based machine translation. Instead, we made them automatically using an SMT tool. In the experiments, we obtained high-quality translation sentences under certain conditions. The proposed method was especially effective in the human evaluation in the A-rank classification. For example, for the simple sentences, it obtained 9 correct sentences and 0 incorrect sentences than the baseline(Moses).
In the future, we will make a program using phrase-based pattern machine translation in order to increase the number of translations in the A-rank.