Conceptual design of an ontological approach to personalising GUISET portal
University of Zululand
Searching and finding specific information from web-based information systems has become tedious and time consuming due to information overload on the web. The current adaptation and personalisation techniques employed are gradually becoming inadequate, as the information available on the web grows exponentially. Measures were taken towards improving the current solutions. The Generic Adaptation Framework (GAF) has been proposed as a standard for building personalised web-based systems. It identifies the standard adaptation components and how adaptation process should be done. The components of the GAF overlay model are only conceptualised to interact at syntax level, which limits interoperability between the components and subsequently the quality of the personalisation results (recommendations). This work proposes semantic interaction of the GAF modelling components, which will not only improve the interaction, but also enhance the overall adaptation process. To enhance personalisation, this work introduced ontologies to model the knowledge about the components and ontology mapping to support meaningful inter-component interaction. The proposed solution, Ontology-based GAF (O-GAF), was applied in a recommender system and compared with the classical GAF recommender system. The experiments carried out showed the O-GAF performs better than the classical GAF in terms of its recommendation accuracy.
A thesis submitted to the Faculty of Science and Agriculture in fulfilment of the requirements for the Degree of Master of Science in Computer Science in the Department of Computer Science at the University of Zululand, South Africa, 2014
web based information systems --ontology --GUISET portal