Ontology Matching Supported by Query Answering in a P2P System

International audienceIn this article we propose methods to automate the extraction of alignments or mappings between ontologies by using query answering in the peer data management system (PDMS) SomeRDFS which uses a data model based on RDF(S). Query answering is composed of a rewriting and an evaluation phase. We show that query answering provides information that offer automated support for discovering new mappings. It is used to generate either mapping shortcuts corresponding to mapping compositions or mappings which can not be inferred from the network but yet relevant. The strategy followed by a peer and filtering criteria defned by the administrator of a peer are used to select the most relevant mappings to be represented