Experience-based Quality Assessment of Distributed Knowledge Graphs

Abstract

This paper introduces an experience-based approach for the evaluation of distributed knowledge graphs. The quality assessment becomes more important in recent days, since distributed knowledge emerges rapidly in different application areas. The paper reports the domain of industrial configuration and production, where distributed knowledge bases have been maintained manually over decades. We describe the configuration ontology COOM and show how standard technologies can be used to query experience-based anomalies. A selection of anomalies is discussed.