On the Impact of Ontological Commitment

Abstract

Ontological commitment, or the agreement to have your applications and users conform to a common domain understanding as encapsulated in one or more shared ontologies, is a noble goal and essential for open agent systems. Our experiences building ontology-based agent systems in multiple domains have shown us that the intention for a new application to locate and conform to some existing ontology or ontologies within its domain has many impediments to its success. For instance, the goals of the designer of a domain ontology include developing a complete and comprehensive domain description; however, the application developer may only require a small fragment of that ontology. Multiple applications that conform to the ontology may, in fact, use completely orthogonal fragments of the ontology, and not be able to interact at all. Users may insist on importing into the ontology sets of terms that are neither logically consistent nor easily modelable.

With these issues in mind, we propose here some guidelines for ontology development and evolution that should facilitate ontology reuse. These guidelines could underpin a usage model for ontologies; one that enables the application designer to reuse ontological concepts from multiple ontologies in a more flexible manner, while retaining the essentially good properties of ontology sharing and reuse. These guidelines affect both the design and use of ontology-based applications, as well as the way applications advertise themselves to other agents with which they may interoperate.