Dealing with Incompatibilities During a Knowledge Bases Fusion Process

Fabien Amarger (1) , Jean-Pierre Chanet (2) , Ollivier Haemmerlé (1, 3) , Nathalie Jane Hernandez (1, 3) , Catherine Roussey (2)

1 IRIT-MELODI - MEthodes et ingénierie des Langues, des Ontologies et du DIscours
2 UR TSCF - Technologies et systèmes d'information pour les agrosystèmes
3 UT2J - Université Toulouse - Jean Jaurès

More and more data sets are published on the linked open data. Reusing these data is a challenging task as for a given domain, several data sets built for specific usage may exist. In this article we present an approach for existing knowledge bases fusion by taking into account incompatibilities that may appear in their representations. Equivalence mappings established by an alignment tool are considered in order to generate a subset of compatible candidates. The approach has been evaluated by domain experts on datasets dealing with agriculture.

Mots clés: Incompatibilities; Knowledge base fusion; Knowledge acquisition

Domaines: Intelligence artificielle [cs.AI]; Apprentissage [cs.LG]; Logique en informatique [cs.LO]; Informatique et langage [cs.CL]

Fabien Amarger, Jean-Pierre Chanet, Ollivier Haemmerlé, Nathalie Jane Hernandez, Catherine Roussey. Dealing with Incompatibilities During a Knowledge Bases Fusion Process. International Conference on Conceptual Structures (ICCS 2016), Jul 2016, Annecy, France. pp. 252-260. ⟨hal-01709190⟩