A multi-criteria approach for automatic ontology recommendation using collective knowledge

Abstract

Nowadays, ontologies are considered an important tool for knowledge structuring and reusing, especially in domains in which the proper organization and processing of information are critical issues (e.g. biomedicine). In these domains, the number of available ontologies has grown rapidly during the last years. This is very positive because it enables a more effective (or more intelligent) knowledge management. However, it raises a new problem: what ontology should be used for a given task? In this work, an approach for the automatic recommendation of ontologies is proposed. This approach is based on measuring the adequacy of an ontology to a given context according to three independent criteria: (i) the extent to which the ontology covers the context, (ii) the semantic richness of the ontology in the context, and (iii) the popularity of the ontology in the Web 2.0. Results show the importance of using collective knowledge in the fields of ontology evaluation and recommendation. © Springer-Verlag Berlin Heidelberg 2012.