Evolution-based refinement of cross-language ontology alignments

Ontology alignment plays a key role for information interconnectivity between computational systems relying on ontologies described in different natural languages. Existing approaches for ontology matching usually provide equivalent type of relation in the generated mappings. In this article, we propose a refinement technique to enable the update of the semantic type of the mapping such as “is-a”, “part-of”, etc. Our approach relies on information from the ontology evolution to apply refinement actions. We formalize the refinement actions and procedures, as well as apply the proposal in application scenarios.