Chinese Word Sense Disambiguation Based On Semantic

The phenomenon of polysemy is caused by the flexibility of Chinese and itis a difficult problem in the Natural Language Processing. Many problems ofNatural Language Processing are ultimately classified as sovling the problems ofambiguous of words. It is a primary problem to solve for WSD that how to makethe computer to be able to deal with the phenomenon of polysemy according tothe specific context. Word sense disambiguation has an important role in manyapplication of Natural Language Processing.The supervised WSD method based on the statistical learning theory is themain research content of this paper. This method is one of mainstream currentmethods in the field of word sense disambiguation.This method preferably hasscalability, flexibility and can adapt to the changes and development oflanguage.This method has been widely applied in the field of domestic andinternational related research. The main content of this paper is consisted by thefollowing parts:Firstly, in this paper, the method of guiding significancelist for the field ofWSD is list and the evaluation method of the accuracy of WSD is introduced. Theunsolved issue of influence on the reserch of WSD is elaborated.Secondly, the method of feature extraction is reserched, including theprocess of corpus analysis, feature selection and feature extraction. The methodof feature extraction based on semantics is deeply, consisting principally of threekinds of Extraction of different semantic information. It includes single layer,three layers of semantic information and the word segmentation information. Andthe characteristic vector set is established by three different feature extractionmethod. Bayesian word sense disambiguation classifier based on semanticinformation is structured by using three kinds of vector of feature extraction. Andthe word sense disambiguation performance is test by using contrast experiment. Finally, the extended application of WSD in other branches of naturallanguage processing is made example. The module of word sense disambiguationis inserted into the machine translation system for raising the accuracy ofmachine translation in application. Although the study is not completed well, itprovides several realistic meaning for the word sense disambiguation in practicalapplication.