Research On The Construction Of Mongolian Emotional Thesaurus Based On The Corpus Of "Da·Nachvgdorji's Works"

At present,the research on information processing of Mongolian sentiment words,it has not been systematically studied,especially based on corpus to construct sentiment word database and analyze text sentiment.The sentiment dictionary is an important tool for judging the sentiment tendency of words and texts.From the perspective of language resource construction,the process of Mongolian sentiment word resource construction,compared with English and Chinese,is very lagging behind and couldn't satisfy users need.This thesis,based on the corpus of Da·Nachvgdorji's Works,try to build the Mongolian sentiment word database and conduct statistical research and analysis.According to the method and theory of corpus linguistics,first work have done in this thesis is,the corpus of “Da·Nachvgdorji's Works” created and marked,which provides basic resources for the construction of the "Mongolian Emotional Thesaurus".Second of all,combined with results of previous studies—the knowledge of mongolian semantic classification system and "Mongolian Semantic Information Dictionary" to preliminarily build an emotional Thesaurus.Then,by matching the thesaurus with the corpus of " Da·Nachvgdorji's work",the "Mongolian Emotional Thesaurus " was constructed and the scale of the thesaurus was expanded.After that,it's analyzed statistically that the distribution of sentiment words in different genre and different parts of speech in “The Works of Da·Nachvgdorji”.At the end of the thesis,it developed corpus and sentiment thesaurus retrieval software,and successfully accomplished the analysis experiment of the sentence sentiment tendency.This thesis consists of an introduction and three main contents.The introduction of this thesis described the basis and significance of the topic selection,research status,research methods and corpus sources,research content and specific research steps.The first chapter of this thesis discussed the basic contents of "The Works of Da·Nachvgdorji ",such as the version overview,corpus overview,corpus development and labeling methods.The second chapter of this thesis,first of all,introduced the Mongolian Semantic Classification System and the process of collecting the information which we needed from “the Mongolian Semantic Information Dictionary”.Then pointed out that used the semantic classification information of the dictionary library to build the "Mongolian emotional thesaurus",and the realization method and process of expanding the thesaurus based on the corpus of " Da·Nachvgdorji's work".At the end,statistically analyzed the distribution characteristics of emotional words in different genres and different parts of speech of "The Works of Da·Nachvg dorji ".The third chapter of this thesis,made graphs to introduce the corpus of "The Works of Da·Nachvgdorji ",the principles and functions of sentiment thesaurus retrieval software.To illustrate the realization of the experiment of sentiment tendency judgment of sentence text and display the application interface.The conclusion part of this thesis,summarized the whole thesis,and discussed some shortcomings of the thesis.