Instance Matching for Ontology Population ? Paper type : FULL PAPER

In the context of ontology evolution, ontology population is the activity of acquiring new semantic descriptions of data extracted from heterogeneous data sources. To this end, the capability of comparing several instances extracted from different sources is crucial. In this paper, we focus on the problem of instance matching and its role for ontology population. Moreover, we present the instance matching systems HMatch(I) which has been developed in the framework of the BOEMIE research project.