A utility-based semantic recommender for technology-enhanced learning

In this paper, we present the design of a Knowledge-based recommender system for Technology Enhanced Learning based on Semantic Web Technologies. It uses a knowledge model for representing the current state of the learner, pedagogical strategies, and learning objects. To create a learner model, the learners' activity and progress is tracked and higher-level learner features (i.e., Didactical Factors) are extracted. For a given learner state and set of pedagogical rules, the Recommendation Engine infers learning objects that lie on the learners personalized learning path. Furthermore, utility functions are used to compute a relevancy score for the best-fit learning objects. We describe the semantic-based recommendation approach on a conceptual level, discuss the strengths and weaknesses on the recommender framework and discuss future research.