A New Method of Computing Chinese Word Similarity Based on Statistics | IEEE Conference Publication | IEEE Xplore