Study On The Method Of Tibetan-Chinese Machine Translation Based On Bilingual Aligned Sentence Patterns Library

Machine translation is a very popular and challenging research content of natural language processing field.Since machine translation method was proposed in the 1930s until now the syntactic information is incorporated into the machine translation,scholars at home and abroad designs a lot of formal model and algorithm of machine translation which make a positive contribution to machine translation of natural language processing research and make it become more mature.However the machine translation is a comprehensive strong research topic,so far,it is still a worldwide problem.The first chapter describes the research significance and research status of machine translation at home and abroad.The second chapter outlines machine translation method based on rules and based on statistical.The third chapter straight into the theme,firstly made a classification of Tibetan language patterns and contrasted the Tibetan-Chinese sentence patterns,and on the basis designed Tibetan-Chinese dictionary for machine translation,built a size of nearly 500 patterns of Tibetan-Chinese aligned pattern library which is the syntactic information incorporated into the machine translation.This pattern library do not include Tibetan-Chinese bilingual alignment syntax tree.It reflects the difference between Chinese and Tibetan sentence pattern by word order and by adding or deleting part of speech.On the premise of bilingual dictionary and aligned sentence pattern library,this paper propose the algorithm of machine translation based on aligned bilingual sentence pattern library whichis the core content of this paper.Finally,several typical example expounds the implementation process of this method in detail,and put forward a solution to some questions in detailsThe fourth chapter firstly reviews research process of machine translation method based on the bilingual aligned patterns library and puts forward the research content needed to complement and perfect in future;Secondly starting from the overall machine translation to look back the forward process and look into the future development trends.