A Study On Mongolian Beginning、Ending And Expositive-valence Automatic Recognition

Summed up the rule of Beginningx Ending and Expositive-valence automatic recognition and successfully completed is the essential factors of the computer generated a correct statement. In this paper, we selected the sentences of Beginning> Ending and Expositive-valence from "the modern Mongolian one million word level corpus" and the 2011,2012 and 2013 the "hua de yuan ye" two million word level corpus and conducted a study.This paper consists of five parts:introduction, the first chapter, the second chapter, the third chapter and the conclusion.The introduction part explains the reason of choosing topic, the research situation, the theory and method of the research, the research steps and the source of the data.The first chapter, mainly introduce the situation of the selected sentence from "the modern Mongolian one million word level corpus" and the 2011,2012 and 2013 the "hua de yuan ye" two million word level corpus, and screened out the sentence for lexical, syntactic and semantic analysis.The second chapter analyzes the Beginning sentence、Ending sentence and Expositive sentence’s semantic collocation.The third chapter, based on the Beginning sentence、Ending sentence and Expositive sentence’s characteristics analysis, summed up the automatic recognition rules of Beginning、Ending and Expositive-valence and made formal description.The conclusion part summarizes the content of the whole thesis, and puts forward some thoughts on the future work.