Study And Implementation Of An Example-based Chinese Uyghur And Uyghur Chinese Bilateral Machine Translation System

Machine translation is a task of translating texts from one naturallanguage to another by computer, such as from Uyghur language toChinese or from Chinese to Uyghur. In the different strategies of machinetranslation, Example-based machine translation method avoids thecomplex, in-depth syntax and semantic analysis. And it has the advantagessuch as high translation accuracy, fast speed, and easiness to obtaintranslation knowledge.The paper first reviews the history of development of machine translation,simply presents various methods of machine translation, and describes theprinciple of example-based machine translation method, its systemarchitecture, related issues and the advantages of this method. We mainlyfocus on natural language processing on agglutinative language, and itscurrent statues of machine translation and analyze the method ofconstructing bilingual example corpus and computing sentence similarityfor Uyghur and Chinese sentence, and combing the similar sentences.And we propose designing method of Example-based Chinese Uyghur andUyghur Chinese Bilateral Machine Translation System and implemented it.This paper deals with an exploratory research on flowing aspect of system:how to construct Chinese Uyghur bilingual example corpus; method ofbuilding Uyghur inverted index, method of computing Uyghur sentencesimilarity; on this bases, design and implement Example-based ChineseUyghur and Uyghur Chinese Bilateral Machine Translation System. The result on the small scale of experiment shows that the translation quality onthe input sentences which are from similar corpus is high while the qualityon the sentences with lack of similarity with corpus is low. And there isstill space of improvement on this system in the future.