The Research Of Uygur And Chinese Machine Translation System Based On The Security Field

The rhythm of social and economic life develop more and more quickly, the technical change is change rapidly. People use to computer and dependence are growing day by day. How to let people use computer technology more convenient to become experts and scholars in the fields of computer are the urgent problems to be solved. In this paper, the field of artificial intelligence in computer technology from the Machine Translation starting, focusing on Xinjiang this fertile land, severe anti-terrorism situation, make a study of safety in the field of Uygur and Han Machine Translation based on. At present, the statistical Machine Translation method based on phrase model is most adopted system. Although Machine Translation system based on phrase model has a good performance of the translation, but its use in the treatment of long distance reordering has inborn deficiency. So the research object of this article is Machine Translation system model based on hierarchical phrase. And through the corpus screening technology and reordering method of embedding, make the version the final quality of the Uygur and Han Machine Translation improve.First of all, this article on the Machine Translation makes the global introduction. Mainly on the method Machine Translation birth to the development of history, mainly used in the Machine Translation, and based on statistical machine translation principles and characteristics. Statistical machine translation from the use of different model, this paper are introduced based on the word translation model, the translation model based on the phrase based translation model and the hierarchical phrase based translation model and syntax. Next, the paper of the Uygur and Han Machine Translation are reviewed in this paper, introduces the realization of the translation system security field of Han and Uygur machine based on the general framework and some of the difficulties faced by the establishment of the system.Secondly, in order to improve the quality of translation system. This paper presents an integrated method of screening model and bilingual language model perplexity of the corpus IBM1 corpus evaluation improved, for screening high quality bilingual corpus based on security domain. Screening method in explaining the proposed before the importance of corpus, this paper simply introduces some other corpus and corpus selection screening method in the whole Machine Translation field. Finally, this paper used the performance experiment proved that this method can improve the data screening Machine Translation effective.Finally, this paper also proposes a combined pretreatment, decoding and post processing method of hierarchical phrase based on the adjusted model order. In this paper we propose a method to explain the ordering before adjusting method of order at present common to use objects for the categories in detail. This paper focuses on some phrase reordering method model based on these methods, and why in performance has failed to go beyond the hierarchical phrase based translation model gives a reasonable explanation. Next, ordering method in this paper and used with the experimental demonstration of the proposed is feasible and effective.In the last part of the article, the author has carried on the summary to the full text, and presented his view of the future development in Machine Translation. The author believe that Machine Translation which now seems a high technology will be used more comprehensive in the near future.