Adaptive Graphical User Interface based on User Profile and Ontologies to Recommend Learning Objects

Abstract

A graphical interface model that adapts itself to the user cognitive preferences and that is able to recommend learning objects (LOs) is proposed. The proposal considers the Methontology methodology, which allows representing the knowledge of the domain based on ontologies. In addition, the Prometheus methodology is used for designing the LOs multi-agent recommendation system. The adaptive functionality consists of distributing the panels that make up the interface according to the particular user interacting with the system. The panel distribution is performed based on the characteristics and preferences of each user, which are stored in the user profile. The validation of the proposed user interface model was done by applying the following metrics: 1) system performance, (2) user experience, (3) usability (4) overload persistence, among others. The results show the benefits of using intelligent agents, ontologies, and user profiles to construct adaptive user interfaces.