Grammar Inference and Statistical Machine Translation

NLP researchers face a dilemma: on one side, it is unarguably accepted that languages have internal structure rather than strings of words. On the other side, they nd it very di cult and expensive to write grammars that have good coverage of language structures. Statistical machine translation tries to cope with this problem by ignoring language structures and using a statistical models to depict the translation process. Most of the translation models are word-based. While the approach has achieved surprisingly good performance comparable to the best commercial systems, many questions remain in the machine translation community. Can the statistical word-based translation still perform well on language pairs with radically di erent linguistic structures? How would it function with less training data or with spoken languages? The thesis work investigated these questions. In summary, word-based alignment model is a major cause of errors in German-English statistical spoken language translation. To account for this problem, a structure-based alignment model is introduced. This new model takes advantages of a bilingual grammar inference algorithm, which can automatically acquire shallow phrase structures used by the model. The structure-based model can directly depict the structure di erence between English and German spoken languages. It also results in focused learning of word alignment, therefore it can alleviate the sparse data problem. The structurebased model achieved 11 percent error reduction over the state-of-the-art statistical machine translation models.