On the detection of inconsistencies in RDF data sets and their correction at ontological level

The rapid development of Linked Data leads to a proliferation of errors in pub- lished data, primarily related to inconsistencies between data instances and their related ontologies. This problem alters the reliability of Semantic Web applications when they involve the analysis or the exploitation of heterogeneous rdf data sets. We focus in this article on a way to correct inconsistencies caused by the domain and the range of a property. We present an algorithm to identify the source of these inconsistencies in the ontology, and to provides guidelines to correct or improve the ontology. The localization of the inconsistencies is based on a quantitative comparison between the classes of domains and ranges defined in the ontology and the ones built by the exhaustive analysis of instances used as subject or object in the properties. We show the usefulness of this method on a case study involving dbpedia: we use our approach to diagnose and correct common inconsistencies. Another experiment conducted on a large data set generated by sp2bench validates the scalability of the proposed algorithm.