HowNet-Based Word Semantic Automatic Classification System
Ruzhan Lu · Jisuanji fangzhen · 2004
Word Semantic Classification is a key part of modern Chinese semantic research. This paper presents hownet-based word semantic automatic classification system, which proposes a new idea of word semantic classification. Based on hownet,we develop a coarse classification module which classifies about sixty thousand Chinese common words supplied by hownet to gain 1420 word semantic equivalent classes and has achieved practical result.On the basis of the above work,we develop a detailed classification module,which reclassifies the relative less precise 1420 word classes in detail.Experiments on the system show that it works very well in natural language processing and linguistics research.