Abstract: Chinese terminology extraction is a fundamental issue in Chinese text information processing. It has been applied in many other fields, such as Chinese text indexing and retrieval, text mining, ontology construction, and latent semantic analysis. Based on mutual information and information entropy theory, we proposed a semi-automatic Chinese terminology extraction method by statistical technology, and experimentally verified this method using internet news topic data. Our results show that proposed method can effectively support Chinese term extraction.

Key words: mutual information, information entropy, Chinese terminology extraction