No abstract

Purpose -Aims to present the ontology engineering methodology DILIGENT, a methodology focussing on the evolution of ontologies instead of the initial design, thus recognizing that knowledge is a tangible and moving target.Design/methodology/approach -First describes the methodology as a whole, then detailing one of the five main steps of DILIGENT. The second part describes case studies, either already performed or planned, and what we learned (or expect to learn) from them.Findings -With the case studies it was discovered the strengths and weaknesses of DILIGENT. During the evolution of ontologies, arguments need to be exchanged about the suggested changes. Identifies those kind of arguments which work best for the discussion of ontology changes.Research implications -DILIGENT recognizes ontology engineering methodologies like OnToKnowledge or Methontology as proven useful for the initial design, but expands them with its strong focus on the user-centric further development of the ontology and the provided integration of automatic agents in the process of ontology evolution.Practical implications -With DILIGENT the experience distilled from a number of case studies and this offers the knowledge manager a methodology to work in an ever-changing environment.Originality/value -DILIGENT is the first methodology to put focus not on the initial development of the ontology, but on the user and his usage of the ontology, and on the changes introduced by the user. We take the user's own view seriously and enable feedback towards the evolution of the ontology, stressing the ontology's role as a shared conceptualisation.

No abstract

Abstract.A prerequisite to the success of the Semantic Web are shared ontologies which enable the seamless exchange of information between different parties. Engineering a shared ontology is a social process. Since its participants have slightly different views on the world, a harmonization effort requires discussing the resulting ontology. During the discussion, participants exchange arguments which may support or object to certain ontology engineering decisions. Experience from software engineering shows that tracking exchanged arguments can help users at a later stage to better understand the assumptions underlying the design decisions. Furthermore, as the constructed ontology becomes larger, ontology engineers might argue in a contradictory way without knowing so. In this paper we present an ontology which formalizes the main concepts which are used in an DILIGENT ontology engineering discussion and thus enables tracking arguments and allows for inconsistency detection. We provide an example which is drawn from experiments in an ontology engineering process to construct an ontology for knowledge management in our institute. Having constructed the ontology we also show how automated ontology learning algorithms could be taken as participants in the OE discussion. Hence, we enable the integration of manual, semi-automatic and automatic ontology creation approaches.

No abstract