Towards On the go matching of linked open data ontologies
Isabel F. Cruz, Matteo Palmonari, Federico Caimi, Cosmin Stroe · BOA (University of Milano-Bicocca) · 2011
Creating links between the schemas of published datasets is a key part of the Linked Open Data (LOD) paradigm. The ability to discover these links the go requires for ontology matching techniques to achieve sufficient precision and recall within acceptable execution times. In this paper we add two methods to the AgreementMaker ontology matching system (but that could be used together with other systems). A first method consists of discovering a set of high-quality equivalence mappings based on concept similarity and of inferring subclass mappings from that set. A second method is based on the adoption of background terminology that serves as a mediator. Sets of synonyms and a hierarchy of concepts are taken into account, as well as subclass relationships involving external concepts that are imported into the ontologies being matched. Experimental results show that our approach has good performance and improves precision on most matching tasks when compared with a leading LOD approach.