Abstract
Building rich axiomatic ontologies automatically is a step towards the realization of the Semantic Web. In this paper, we describe an automatic approach to extract complex classes’ axioms from Wikipedia definitions based on recurring syntactic structures. The objective is to enrich DBpedia concept descriptions with formal definitions. We leverage RDF to build a sentence representation and SPARQL to model patterns and their transformations, thus easing the querying of syntactic structures and the reusability of the extracted patterns. Our preliminary evaluation shows that we obtain satisfying results, which will be further improved.
Cite
CITATION STYLE
Haidar-Ahmad, L., Zouaq, A., & Gagnon, M. (2016). Automatic extraction of axioms from wikipedia using SPARQL. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9989 LNCS, pp. 60–64). Springer Verlag. https://doi.org/10.1007/978-3-319-47602-5_13
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
