Obtaining information content of concepts: An overview

Semantic similarity measure between concepts is a generic issue for many applications of computational linguistics and artificial intelligence. Information Content (IC) of concept is an important dimension of accessing semantic similarity between two concepts. Recently it has attracted great concern and become a hot topic. The paper gives a general overview of usage of information content in semantic similarity computing and then focuses on how to obtain information content. It reviews and analyses state-of-art IC models, including Corpus-dependent and Corpus-independent IC approach. Hyponym-based IC Model, leaves-based IC Model, concept's topology structure based IC Model and relation-based IC Model are discussed respectively in detail. The important related issues are described. Finally further research is outlined for the improvement of IC.