The computation of concept similarity based on PSO algorithm
Jia Liu, Huayu Wang · 2010
Ontology heterogeneity restricts the application of ontology. The key of ontology mapping is to compute concept similarity. At present, the method of similarity computation is imperfect. And the computation quantity is high. In order to decrease computation quantity and completely compute concept similarity, the influences of relations and attributes is considered and use candidate set. In this paper, a computation method of similarity based on PSO algorithm is proposed. At first, the semantic similarity and descriptive similarity of concepts are computed. Then the attribute priority is specified by PSO algorithm. At last, the attribute with high priority is chosen according to the user need. Take two ontologies as example and compute concept similarity. Experiments results show that this calculative method can improve the precision of similarity and reduce the calculated quantity.