Research Of Word Sense Disambiguation Based On Word-sense Category Extending

Word Sense Disambiguation(WSD) is a key research project in the Natural Language Process (NLP)and many other field in the NLP. WSD has an important meaning both in theory and practice. Recently WSD has become an important research task both in the computalitonal linguistics and NLP.This article intruduce the main research method and knowledge resource that used in the WSD. An new method is bring out in this article to resolve the common problem we faced in WSD.Recent years the compute-speed of the computer become faster and faster and the meomory capacity become larger and larger. So the method based on statistic develop rapidly. Dictionary reseource and corpus are the main resource in the WSD which based on statictic. Supervised Disambiguation need a corpus which provide word sense inofrmation to be trained. In order to inhence the accuracy of the WSD the extended dictionary is used in this article.Sparse Source Data and little probability word-sense have always been a tough problem in the WSD field. In order to solve the heavy work in artifical corpus and the Sparse Source Data, this article use an Word-sense Category Extending to extend the polysemous word and indicate word. This method can improve the efficiency of the current corpus. The experiment has improve that this method which can settle the problem of Sparse Source Data and little probability word-sense is valid.This measure improve the accuracy of the WSD and provide some reference to WSD.