Découverte de correspondances sémantiques entre ressources hétérogènes dans un environnement coopératif

Authors

Abstract

In collaborative work experts are encouraged to use semantic resources. These resources are generally represented with the aid of concepts and relationships and can be modelled in the form of ontologies. Each expert in a field of application often uses one or more semantic resources, and may also have access to other semantic resources made available by her/his partners. However, these resources are heterogeneous at different levels making difficult syntactic, structural and semantic interoperability. In this work, we emphasize semantic interoperability. We propose a methodology to help semiautomatic discovery of semantic mappings between concepts of different semantic resources. To accomplish this work, we transform these semantic resources in OWL-DL to use the services of inference engine based on description logics. We also propose a process for creating equivalence for relations between different semantic resources, called unification of relations; it aims to discover other semantic mapping. The results of this work were tested by a prototype. The implementation was carried out with resources from within the industry European construction sector. To assist in the evaluation of semantic mappings, we also propose a domain knowledge modelling based on fuzzy logics.

Subject

IS Research

PhD Thesis

Découverte de correspondances sémantiques entre ressources hétérogènes dans un environnement coopératif, December 2007