Domain Independent Learning of Ontology Mappings
Floris Wiesman, Nico Roos · Research Publications (Maastricht University) · 2004
This paper proposes a domain independent method for handling interoperability problems by learning a mapping between ontologies. The learning method is based on exchanging instances of concepts that are defined in the ontologies. The method starts with identifying pairs of instances of concepts denoting the same entity in the world using information retrieval techniques, followed by proposing and evaluating mappings between the ontologies using the pairs of instances. For each step of this method, the likelihood that a decision is correct is taken into account. Important benefits of the method are that (a) no domain knowledge is required, and (b) the structures of ontologies between which a mapping must be established, play no role.