Context Based Ontology Matching using Gold Standard Model and C Metric Analysis

Abstract

The motivation behind current study is reported slowing down in speed of newlineimprovement in the field of ontology matching. Semantic heterogeneity is biggest newlinechallenge in semantic web. Ontology matching is a solution to the semantic newlineheterogeneity problem. It finds correspondences between semantically related entities newlineof ontologies. Ontology matching can be performed at various levels, criteria, and newlineenvironments leading to different kind of techniques. It may be done either locally at newlinethe element level or globally at the structure level of ontologies. Matching criterion newlinecould be semantic, syntactic, terminological, structural and extensional based. newlineMatching environment could be either context or content based. newlineOntology matching using background knowledge improves match result. Therefore newlinechallenges in ontology matching using contextual knowledge are identified. To find newlinesolutions for these challenges, relationship among various matching levels, criteria and newlineenvironments are investigated. Model classification of ontology matching techniques newlineis reviewed and a gold standard model for an ontology matching system in terms of a newlinelayered view of matching levels, criteria and environments is proposed. Based on this newlinemodel and model classification, a fresh classification of ontology matching systems is newlineproposed. An appropriate metric is also proposed to measure the coverage of ontology newlinematching parameters by a class of ontology matching systems as per proposed model. newlineA survey of ontology matching systems reported recently in literature is done and newlinethese systems are classified according to proposed classification. The data collected in newlinesurvey is analyzed as per proposed metric. It is found that semantic systems fare much newlinebetter with respect to proposed metric. It is followed by extensional and syntactic newlinesystems in that order. These results support proposed gold standard model. newlineHence, context-based semantic system (taxonomy structure) is proposed for our newlineontology matching framework. The framework is designed and implemented and its newline