Conclusion

Abstract

After reviewing previous studies, proposing hypotheses and research questions, introducing several useful lexical knowledge bases and corpora, explaining data collection , and discussing the main research approaches—the character similarity clustering analysis and the concept similarity clustering analysis of the corpus-based and computational approach and the off-line multiple-choice tasks of the experimental evaluations—I maintain that my approach for sense prediction in this study is both powerful and valuable. Summary and discussion as well as contributions of this work will be discussed before closing with a discussion on related future work.