Bridging the multilingualism gap in ontology alignment

Abstract

In the semantic web context, research in ontology alignment has focused primarily on the development of monolingual matching algorithms. As more and more resources become available in more than one natural language, novel algorithms are required. The latter are capable of aligning ontologies which share more than one language. Indeed, translation techniques are often used as an intermediate step to translate ontology labels. This treatment essentially removes the natural language barrier in the alignment environment. To meet this challenge, we propose in this paper a new cross-lingual alignment method, designed to take into account the goal of dealing with ontologies expressed in different natural languages. The proposed method implements a strategy of a direct alignment based on the use of external resources. The results of performed tests with our method on different pairs of ontologies and according to several scenarios show their effectiveness.