Protégé based environment for DL knowledge base structural analysis

Abstract

Structural analysis of knowledge bases improves process of obtaining additional hidden information and provides new means of information discovery. In this paper we propose a method of structural analysis and characterise in details developed toolkit as a Protégé extension. The aim of method itself is to identify hidden relationships using structural weighted graph analysis, applied for both terminology and instance base. In order to evaluate new relationships between instances, we have chosen a set of measures that consider the analysis of both vertices and links, thus providing new semantic information for identified associations. Developed environment concludes both approaches (logic based and graph based) in order to apply reasoning techniques, graph based algorithms and measures for inferring important analytical information. The scope of paper provides also application areas of such method in terms of data mining techniques for crisis identification and multi-criteria data analysis. © 2011 Springer-Verlag Berlin Heidelberg.