Ontological knowledge management through hybrid unsupervised clustering techniques

Abstract

In the Semantic Web, ontology plays a prominent role to actualize knowledge sharing and reuse among distributed knowledge sources. Intelligently managing ontological knowledge (classes, properties and instances) enables efficacious ontological interoperability. In this paper, we present a hybrid unsupervised clustering model, which comprises of Formal Concept Analysis, Self-Organizing Map and K-Means for managing ontological knowledge, and lexical matching based on Levenshtein edit distance for retrieving knowledge. The ontological knowledge management framework supports the tasks of adding a new ontological concept, updating and editing an existing ontological concept and querying ontological concepts to facilitate knowledge retrieval through conceptual clustering, cluster-based identification and concept-based query. The framework can be used to facilitate ontology reuse and ontological concept visualization and navigation in concept lattice form through the formal context space. © 2008 Springer-Verlag Berlin Heidelberg.