Fuzzy semantic similarity in linked data using wikipedia infobox

摘要

The problem of semantic similarity assessment arises in several applications, for example, knowledge management, information integration, and information discovery. In this article, we present a new method that evaluates similarity between entities represented by Resource Description Framework (RDF) triples introduced in the context of the Semantic Web. At the beginning, our approach identifies and groups properties according to their importance. It is done via exploiting the information presented in Wikipedia infoboxes. Then semantic similarity corresponding to each group is calculated using both the schema (ontology classes and properties) and RDF links discovered from different datasets (due to the open and distributed nature of data). Finally, the calculated similarity measures for all groups are aggregated using weights obtained from a specially designed fuzzy membership function. Experimental evaluations confirm the suitability of the proposed method.