Ontology Selection Ranking Model based on Semantic Similarity Approach

Ontologies have provided supports in integrating heterogeneous and distributed information. More and more ontologies and tools have been developed in various domains. However, building ontologies requires much time and effort. Therefore, ontologies need to be shared and reused among users. Specifically, finding the desired ontology from an ontology repository will benefit users. In the past, most of the studies on retrieving and ranking ontologies have mainly focused on lexical level supports. In those cases, it is impossible to find an ontology that includes concepts that users want to use at the semantic level. Most ontology libraries and ontology search engines have not provided semantic matching capability. Retrieving an ontology that users want to use requires a new ontology selection and ranking mechanism based on semantic similarity matching. We propose an ontology selection and ranking model consisting of selection criteria and metrics which are enhanced in semantic matching capabilities. The model we propose presents two novel features different from the previous research models. First, it enhances the ontology selection and ranking method practically and effectively by enabling semantic matching of taxonomy or relational linkage between concepts. Second, it identifies what measures should be used to rank ontologies in the given context and what weight should be assigned to each selection measure.