Revising and managing multiple ontology versions in a possible worlds setting

Abstract

To date, ontology research seems to have come to an equilibrium: there is a wide variety of theories, methods and tools for extracting, representing, storing and browsing ontologies. Furthermore, ontologies have become an integral part of many academic and industrial applications. There are, however, still deep theoretical problems to be resolved. One fundamental problem is ontology evolution. Essentially, changes occur when new knowledge is agreed on by the (not necessarily human) cognitive agents. Each change engages an evolution process from which a new (possibly inconsistent) ontology version may emerge. How will we manage multiple versions of an ontology in a scalable fashion? How will we represent differences and transformations between ontologies, and moreover, can we solve this question independently from any ontology representation model? This paper describes a research project that will investigate all these critical problems in ontology evolution. © Springer-Verlag 2004.