Ontology matching systems take a prominent position in solving semantic heterogeneity problems to facilitate sharing and reuse of ontologies. The process of generating ontology alignments through ontology matching techniques purely lies on how the concepts and relationships are modeled. This paper focuses on designing an ontology matching system in which concepts are modeled based on cognitive units of knowledge comprising of objects, attributes and relationships. The proposed cognitive based ontology matching system(COGOM) identifies semantically related concepts by aggregating the attribute similarity degree, structural similarity degree and semantic conception degree. The similarity computation is adapted from the Tversky psychological model of similarity. The proposed ontology matching system is adaptive in nature because of the cognitive based knowledge expression and the computational overhead of generating alignments is improved by forming quality clusters of semantically correlating concepts thus reducing the concept match space. The precision and recall metrics are used for evaluation of the proposed system using the benchmark data sets of OAEI 2015.
展开▼
