Computation of Conceptual Similarity in Ontology Mapping
Qin Duo-rong · Computer Technology and Development · 2008
Ontology is usually viewed as the sets of concepts,attributes and relations.ontology mapping is the best way to solve ontology heterogeneity.To aim at the current problems of the computation of concept similarity,puts forward an improved approach.Firstly,the most related concepts are filtered out according to the similarity between two concept names so as to reduce the similarity computation,and then an integrated approach of based-concept attribute,based-concept instance,based-concept relation is designed to compute the concept similarity.When calculating attribute similarity,determine the priority of various attributes by computing their entropies,then only compute the attribute similarity through several attributes with relative large entropy values to reduce the amount of computation.