STATISTICAL TRANSLATION APPARATUS BY SEPARATING SYNTACTIC TRANSLATION MODEL FROM LEXICAL TRANSLATION MODEL AND STATISTICAL TRANSLATION METHOD

摘要

The present invention relates to statistical machine translation, which extracts a syntactic structure conversion knowledge and a lexical translation knowledge from parallel corpus, extracts each conversion probability, acquires a generation probability for a target language from a single corpus, And the probability is modeled by a translation model learning machine, and a weighted translation model is modeled. Then, the modeled translation model is applied to the input source sentence in real time, and the decoding process of the syntax structure converter and the lexical converter is performed, Based SMT and syntactic structure-based SMT, and provides a machine translation apparatus and a machine translation method which combine the merits of the present invention to combine the advantages of the existing phrase-based SMT and syntax-based SMT.