Instance matching based on discriminative property values

Abstract

Ontologies in large scale contain a huge number of instances. Linking instances from different data sources that denote the same real-world object present a challenge for Linked Data. For ontological data integration and sharing, especially with rapid development of linked data, automatically instance matching has been become the fundamental issue. In this paper, we propose a new instance matching approach. The proposed method is based on the discriminative property values. Only the similar instances which have similar discriminative property values are selected to compare their descriptive property values. Two types of link will be established between similar instances SameAs and the proposed novel link ViewSameAs. This link is established between instances which have similar discriminative property values and dissimilar descriptive property values. It aims to save the initial result of our matching process.