Ontology Matching for Socio-Cyberphysical Systems: An Approach Based on Background Knowledge

Nowadays there are exist many ontology matching methods mostly based on using context and content in various combinations. The accuracy of existing methods can be increased by using the background knowledge – an external reference knowledge that can facilitate the matching process. The paper proposes an approach for automated matching of elements (fragments) of the ontologies based on the combination of existed ontology matching methods (pattern and context-based), neural network matching and additional control by community-driven matching. All of them are using the background knowledge to get additional information that helps to find more precise correspondence between ontologies concepts. Also, the background knowledge helps to explain result of ontology matching to the user, which is going to check the correspondence manually. In addition, the paper proposes an architecture of ontology matching service that is built based on the presented approach. The service is used for providing semantic interoperability in socio-cyberphysical systems.