Fuzzy Ontologies and Scale-free Networks Analysis

Abstract

In the recent years ontologies have played a major role in knowledge representation, both in the theoretic aspects and in many application domains (e.g., Semantic Web, Semantic Web Services, Information Retrieval Systems). The structure provided by an ontology lets us to semantically reason with the concepts. In this paper, we present a novel kind of concept network based on the evolution of a dynamical fuzzy ontology. A dynamical fuzzy ontology lets us to manage vague and imprecise information. Fuzzy ontologies have been deflned by integrating Fuzzy Set Theory into ontology domain, so that a truth value is assigned to each concept and relation. In particular, we have examined the case where the truth values change in time according to the queries executed on the represented knowledge domain. Empirically we show how the concepts and relations evolve towards a power-law statistical distribution. This distribution is the same that characterizes complex network systems. The fuzzy concept network evolution is analyzed as a new case of a scale-free system. Two e‐ciency measures are evaluated on such a network at difierent evolution stages. A novel information retrieval algorithm using fuzzy concept networks is also proposed.