Word Cluster Based on HNC
Cai Yue-hong · Journal of Zhengzhou University · 2009
To solve the problems of redundant information and computational complexity,an antonym,synonym,near-synonym clustering algorithm is proposed based on HNC.The basic idea of the algorithm is mapping semantic meaning of the words to the system of HNC concept symbols,and thus all the words become a series of symbols.And based on calculating the semantic similarity and semantic distance,the clustering of antonym,synonym and near-synonym is completed on HNC word symbol corpus.