Study And Implementation On Key Techniques For Example Based Machine Translation

In this paper, machine translation of the core of a few, summed up the machine translation system framework, discussed the modeling machine translation factors should be taken into account, a detailed analysis of segmentation, alignment sentence, the translation of knowledge acquisition, translation clips choice and in the machine translation: Generation's role in the process, completed a Case-Based Machine Translation modeling, achieving a case-based machine translation system.1. Research and analysis on the impact of machine translation of several core of summed up the machine translation system framework, in-depth study of the machine translation modeling considerations;2. Studied several common sentence alignment model, a model of integrated sentence alignment. Experiments show that, and the previous sentence alignment method, the model significantly increased the sentence to the correct alignment;3. Put on the basic translation of the fragments, using vocabulary alignment results from the realization of bilingual corpus automatically based on access to basic translation of the fragments. Do not use this method because of parsing knowledge, in any theory can be taken from the Corpus bilingual translation clips, and so they have the language independence.In this paper, based on the example of the complete machine translation modeling, and the achievement of the model. Experimental data indicate that, this paper presents the case based machine translation system can the completion of machine translation, and have automatic access to knowledge translation, language independent features.