Ontology matching by actively propagating user feedbacks through upper ontologies Juego de ontología que propaga activamente evaluaciones de usuarios a través de ontologías superiores

Ontology matching is a complex and largely user-driven process of finding correspondences between entities belonging to different ontologies. Many algorithms have been proposed to automate the matching generation. However, they can’t be fully automated since the user input is required to accept, reject, or create new alignments or matchings. This paper extends on active learning framework for ontology matching, which tries to find the most informative candidate matches to query the user. In our approach the user’s feedback exploits upper ontologies as semantic bridges. Such bridges contribute to the overall matching process while considering the supervised information and its propagation in correcting mistake matchings. In the experimentation our work outperformed the previous version where none upper ontology was used, while it remains as competitive as state of the art ontology matching system.