Application of an Ontology-Based User Model for Content Personalization and Recommendation in Web Portals

A web portal is an application that provides one-point unified access to a multitude of information resources and services residing on the Web and in the corporate intranets. One of the biggest challenges of  today’s web portals is the constantly growing amount of information being provided, which makes it more difficult and time consuming for the user to find relevant content. In order to mitigate this problem, the user should be provided with personalized access to the portal’s content, i.e., the portal should be able to automatically identify the most relevant content (based on the user's individual interests) and display it in such a way that the user can quickly and easily access it.

The main goal of this thesis is to develop a set of mechanisms for personalizing the portal content based on the ontology-based user interest model developed in our earlier work [ 1, 2]. This set should include the following methods: sorting a list of portal documents by user interest, displaying a list of recommended documents based on a given portal section and the user model, and highlighting interesting concepts in the text. The scope of this work includes but is not limited to a literature review in e related research areas (recommender and adaptive systems, web portals, ontologies, and usability), prototyping, usability testing, implementation of personalized portlets in Java and JSP, and deployment in a portal.

Die Arbeit kann selbstverständlich in Englisch oder Deutsch verfasst werden.

Für weitere Informationen melden Sie sich bitte bei Prof. Dr. Birgitta König-Ries oder Fedor Bakalov.