A Peer-review Approach for Ontology Evaluation
Kaustubh Supekar · 2005
The semantic web is intended for knowledge sharing among agents as well as humans [BernersLee et al., 2001]. To achieve this goal, ontologies, which express knowledge in a certain representation as well as in machine interpretable form were introduced, and have grown considerably in number. To have a sustained growth of the semantic web and to have better interoperability between intelligent systems and applications it is highly desirable and is very critical to reuse existing ontologies. Furthermore, ontology engineering is an exceedingly intricate and challenging task requiring specialized design skills as well as comprehensive domain knowledge. Reuse of ontologies present on the web will ease this burden. Users who wish to reuse these ontologies, however, are often confronted with knowledge resources that cover overlapping domains of interests, and that vary in quality of their content. To our knowledge, currently, there are no operational quantitative or qualitative methodologies to assess the quality of ontology content. Consequently, more often than not knowledge engineers will develop their ontologies from scratch rather than reuse existing ones.