Concept IC Model in WordNet Based on Entropy
He Yan, Zili Zhou · Jisuanji gongcheng · 2013
(Abstract) )) )A new Information Content(IC) model is given in this paper which can be used to calculate the IC value of concept in WordNet. The concept of entropy is introduced in the model which considers not only the number of child node of concept and the depth in the tree of taxonomy of WordNet, but also the spatial structure of hyponyms of the concept, the model makes the value of IC of the concept more accurate. Experimental results show that the precision of the semantic similarity algorithms using the IC values computed by the entropy model can be improved.