Management of dynamic knowledge

Purpose: This paper presents a framework for ontology evolution tailored to Digital Libraries, which makes use of two different sources for change detection and propagation, the usage of ontologies by users and the changes of available data. Approach: After presenting the logical architecture of the evolution framework, we first illustrate how to deal with usage-driven changes, that is changes derived from the actual usage of ontologies. Second, we describe the generation of data-driven ontology changes based on the constant flow of documents coming into Digital Libraries. Findings: The proposed framework for ontology ontology evolution, which is currently applied and evaluated in the case studies, significantly reduces the costs of ontology updates and improves the quality of the ontology with respect to the users’ requirements. Research implications: Practical implications: The management of dynamic knowledge is crucial for many knowledge management applications. Our approach for usage-driven and data-driven change discovery not only assures the consistency of ontologies modeling dynamic knowledge, but also reduces the burden of manual ontology engineering. Originality: This paper presents the first approach towards a common framework for ontology evolution based on usage-driven and data-driven change discov-