English Language Statistical Machine Translation Oriented Classification Algorithm

In this paper, we study on the problem of statistical machine translation (SMT) for English language, and the main innovation is that we introduce the SVM classifier to solve this problem. Particularly, the process of statistical machine translation is constructed of two modules, that is, 1) constructing the target-language sentence, and 2) translating from target texts to source texts language. Afterwards, framework of the statistical machine translation system is provided. The main idea of statistical machine translation is in that the sentence with highest probability is selected by optimizing an objective function, which describes the relationships between source language sentences and target language sentences. Finally, experimental results demonstrate that our proposed algorithm outperforms Topic-Based SMT and Rule-Based SMT.