2007 | OriginalPaper | Chapter
Optimal Learning of Ontology Mappings from Human Interactions
Authors : Sumit Sen, Dolphy Fernandes, N. L. Sarda
Published in: On the Move to Meaningful Internet Systems 2007: CoopIS, DOA, ODBASE, GADA, and IS
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Lexical similarity based ontology mappings are useful to obtain semantic translations of database schemas across application domains. Incremental improvement of such mappings can be obtained from human inputs of ontology mapping. Manual mappings are labor intensive and need to be assisted by machine-generated mappings in a semi-automated approach. Heuristics based approaches allow multiple strategies to learn human expertise in concept mappings. Such learning improves the level of automation of the mapping process. We analyze heuristics based Bayesian learning of manual mappings to improve effectiveness of machine-generated mappings. Our results show that human based mappings contribute higher improvement in the machine-generated values of lexical similarity in comparison to those of structural similarity. The optimal weightage for structural similarity learning is inversely proportional to the complexity of given ontology graphs.