Affiliations: Labged Laboratory: Laboratory of Electronic Document Management, Computer Science Department, Badji Mokhtar University Annaba, Annaba, Algeria | [x] Business Information Systems, University College Cork, Cork, Ireland | [y] Information Systems and Operations Management, The Sellinger School of Business and Management, Loyola University Maryland, Baltimore, MD, USA
Abstract: The dynamic domains change over time, in which the modified knowledge in the Ontology that represents the domain has to be considered. In such situation, several versions of the same Ontology can be found with unknown differences between them. Several methods for Ontology's alignment exist, however very few are interested in the comparison between versions. This study implements a method of comparing Ontology's versions. It extracts differences between OWL Ontology's versions. In order to calculate the distance between two Concepts, the method is based on the combination of similarity measures proposed by Levenshtein and OWL's scheme of characteristics. The algorithm used within the method integrates different heuristics matchers for comparing Ontology versions. At the end of the study, a description of Software tool is given. The result of executing this approach on three versions of Cancer disease Ontology is summarized in a change data- base that includes all the existing differences. For the evaluation, two criteria: Recall and Precision are used to validate the model.
Keywords: Ontology versioning, OWL, information retrieval, COMPONTO