A method for creating and using a context knowledge base for ontology integration
2014 (English)In: Information Sciences, ISSN 0020-0255, E-ISSN 1872-6291Article in journal, Editorial material (Refereed) Submitted
Ontology integrations are applied for generating new ontologies used for providing advanced services. Although many algorithms and systems for ontology integration have been proposed, it is still very difficult to achieve semantic and pragmatic ontology matching and integration. To tackle the problems, a method of building and using a context knowledge base is needed to get the context of ontologies used for matching and integration. In the method, a context knowledge base with context rules is developed upon an ontology repository to improve the ontology integration. In the context knowledge base, context rules use Ontology Metadata Vocabulary to describe contexts. The stored ontologies that in the repository satisfy a particular context, are extracted and integrated automatically to form contextual information. In the contextual information, both the consistent ontological definitions and the different perspectives from various ontologies of string-identical entities are used. For new ontology integrations, context rules are searched and triggered according to the context of the ontology integration at hand. The contextual information of the rules is aggregated for improving the new ontology integration. Meta-rules are also automatically built to be able to apply the context rules in a hierarchical relation. Experiments of integrations show that the context knowledge base provides extra contextual information for ontology integration. Moreover, when comparing an integration using the context knowledge base to an ontology integration without the context knowledge base, the results of the contextual ontology integrations are improved. At the same time, it is observed that the quantity and the quality of the stored ontologies, in a repository, determine the quality of the context knowledge base. The better quality and quantity of the context knowledge base, the higher degree of improvement it infers to the contextual ontology integrations.
Place, publisher, year, edition, pages
Elsevier , 2014.
ontology integration, context, knowledge base, rules, contextual ontology intersection
Research subject SRA - ICT
IdentifiersURN: urn:nbn:se:kth:diva-154140OAI: oai:DiVA.org:kth-154140DiVA: diva2:755349
QS 20152014-10-142014-10-142015-04-14Bibliographically approved