A tool for theme identification in RDF graphs

Abstract

An increasing number of RDF datasets is published on the Web. A user willing to use these datasets will first have to explore them in order to determine which information is relevant for his own needs. To facilitate this exploration, we present a system which provides a thematic view of a given RDF dataset, making it easier to target the relevant resources and properties. Our system combines a density-based graph clustering algorithm with semantic clustering criteria in order to identify clusters, each one corresponding to a theme. In this paper, we will give an overview of our approach for theme identification and we will present our system along with a scenario illustrating its main features. © Springer International Publishing Switzerland 2014.