Dalhousie Repository

K-MORPH: Knowledge Morphing via Reconciliation of Contextualized Sub-ontologies

DalSpace/Manakin Repository

Show simple item record

dc.contributor.author Hussain, Syed Sajjad
dc.date.accessioned 2011-04-12T11:47:10Z
dc.date.available 2011-04-12T11:47:10Z
dc.date.issued 2011-04-12
dc.identifier.uri http://hdl.handle.net/10222/13350
dc.description.abstract Knowledge-driven problem solving demands 'complete' knowledge about the domain and its interpretation under different contexts. Knowledge Morphing aims at a context-driven integration of heterogeneous knowledge sources--in order to provide a comprehensive and networked view of all knowledge about a domain-specific problem, pertaining to the context at hand. In this PhD thesis, we have proposed a Semantic Web based framework, K-MORPH, for Knowledge Morphing via Reconciliation of Contextualized Sub-ontologies. In order to realize our K-MORPH framework, we have developed: (i) a sub-ontology extraction method for generating contextualized sub-ontologies from the source ontologies pertinent to the problem-context at hand; (ii) two ontology matching approaches: triple-based ontology matching (TOM) and proof-based ontology matching (POM) for finding both atomic and complex correspondences between two extracted contextualized sub-ontologies; and (iii) our approach for resolving inconsistencies in ontologies by generating minimal inconsistent resolve candidates (MIRCs), where removing any of the MIRCs from the inconsistent ontology results in a maximal consistent sub-ontology. Thus, K-MORPH performs knowledge morphing among ontology-modelled knowledge sources and generates a context-sensitive and comprehensive knowledge-base pertinent to the problem at hand by (a) extracting problem-specific knowledge components from ontology-modelled knowledge sources using our sub-ontology extraction method; (b) aligning and merging the extracted knowledge components using our matching approaches; and (c) repairing inconsistencies in the morphed knowledge by applying our approach for detecting and resolving inconsistencies. We demonstrated the application of our K-MORPH framework in the healthcare domain, where K-MORPH generated a merged ontology for providing a comprehensive therapeutic knowledge-base for Urinary Tract Infections (UTI) by first (i) extracting 20 contextualized sub-ontologies from various UTI ontologies of different healthcare institutions, (ii) aligning and merging the extracted UTI sub-ontologies, and (iii) detecting and resolving inconsistencies in the merged UTI ontology. en_US
dc.language.iso en en_US
dc.subject Knowledge Management en_US
dc.subject Knowledge Integration en_US
dc.subject Semantic Web en_US
dc.subject Ontology Modularization en_US
dc.subject Ontology Matching en_US
dc.subject Ontology Debugging en_US
dc.title K-MORPH: Knowledge Morphing via Reconciliation of Contextualized Sub-ontologies en_US
dc.date.defence 2011-03-29
dc.contributor.department Faculty of Computer Science en_US
dc.contributor.degree Doctor of Philosophy en_US
dc.contributor.external-examiner Dr. Mor Peleg en_US
dc.contributor.graduate-coordinator Dr. Malcom Heywood en_US
dc.contributor.thesis-reader Dr. Michael Shepherd en_US
dc.contributor.thesis-reader Dr. Denis Riordan en_US
dc.contributor.thesis-supervisor Dr. Syed Sibte Raza Abidi en_US
dc.contributor.ethics-approval Not Applicable en_US
dc.contributor.manuscripts Not Applicable en_US
dc.contributor.copyright-release Not Applicable en_US

Files in this item

This item appears in the following Collection(s)

Show simple item record