A Concept Similarity Method in Structural and Semantic Levels

Based on the formal concept analysis, combining the rough set and semantic ontology, this paper proposes a formal concept similarity measurement by improving the original Tversky similarity model. The method used the hierarchical information of concept lattice in the lattice structure and its own semantic meaning to measure the similarity of from the structure information of the concepts and semantic information respectively, which reflect the true similarity of formal concept to some extent.