Ontology-based semantic similarity: A new approach based on analysis of the concept intent

Ontology offers a structured knowledge representation and provides formal interpretation of concepts which can be used in semantic similarity measuring. In this paper, we analyze these ontology-based approaches for semantic similarity computation and propose a new ontology-based measure relying on exploiting intent of the concept. Our measurement combines the idea of two popular semantic similarity calculation approaches: graph-based approaches and feature-based measures. In order to compute the semantic similarity of concepts in ontology, a new algorithm is presented. We compare results obtained by our method with other two typical approaches, the results show that our measurement can distinguish fine differences between concepts and thus has finer granularity.