Effects of the ISIS Recommender System for navigation support in self-organised Learning Networks

Abstract

Presentation of the experimental effects of the ISIS recommender system for self-organized Learning Networks. In this study we applied a recommender system that was based on stereotype filtering algorithm and a domain ontology in an experimental Moodle environment. For the stereotype filtering part we took into account successful completed learning activities by other learners as implicit ratings. The ontology was mainly used to cover the ‘cold-start’ problem. Therefore, we took advantage of the explicit association of domain experts in order to recommend learning activities. The recommendation was based on quantitative information about successful completions of activities by learners with similar preferences.