Semantic Web Datatype Similarity: Towards Better RDF Document Matching

Irvin Dongo (1) , Firas Al-Khalil (2) , Richard Chbeir (1) , Yudith Cardinale (3, 1)

1 LIUPPA - Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour
2 UCC - University College Cork
3 USB - Universidad Simon Bolivar

With the advance of the Semantic Web, the need to integrate and combine data from different sources has increased considerably. Many efforts have focused on RDF document matching. However, they present limited approaches in the context of datatype similarity. This paper addresses the issue of datatype similarity for the Semantic Web as a first step towards a better RDF document matching. We propose a datatype hierarchy, based on W3C's XSD datatype hierarchy, that better captures the subsumption relationship among primitive and derived datatypes. We also propose a new datatype similarity measure, that takes into consideration several aspects related to the new hierarchical relations between compared datatypes. Our experiments show that the new similarity measure, along with the new hierarchy, produces better results (closer to what a human expert would think about the similarity of compared datatypes) than the ones described in the literature. \textcopyright 2017, Springer International Publishing AG.

Domaines: Informatique [cs]

Irvin Dongo, Firas Al-Khalil, Richard Chbeir, Yudith Cardinale. Semantic Web Datatype Similarity: Towards Better RDF Document Matching. Database and Expert Systems Applications - 28th International Conference, DEXA 2017, Lyon, France, August 28-31, 2017, Proceedings, Part I, Aug 2017, Lyon, France. pp.189-205, ⟨10.1007/978-3-319-64468-4_15⟩. ⟨hal-01909095⟩