Personalized Resource Recommendation for a Semantic-enhanced Platform for Information Resources Improving OntoAIMS

Abstract

With the rise of the Semantic Web and freely available web services, the possibilities of searching for information improved. Current information systems, however, often use old-fashioned methods for recommending information resources, based on user models which are based on the behaviour of the user instead of the actual knowledge of the user. An enhanced user model, developed for OntoAIMS, enables improved search methodologies resulting in a personalized resource recommender. This recommender is implemented for the OntoAIMS adaptive information management system which is used as an e-learning environment for testing the results of this recommender. This recommender provides a personalized recommendation system for students, using information in the ontology used for the domain representation and the knowledge of the user stored in an enhanced user model. Furthermore, the recommendation system automatically retrieves resources from the Internet and annotates the resources with available metadata and concepts in the ontology, which is used for improving the recommendation system. Based on an evaluation with real users, the OntoAIMS system is improved with the resource recommender. We present an analysis of this user evaluation and the benefits of a personalized resource recommender. A detailed design and implementation of a personalized resource