A Rule-Based Approach for Adaptive Content Recommendation in a Personalized Learning Environment: An Experimental Analysis

Abstract

Rapidly growing Information and Communication Technologies (ICT) have increased the availability of multimedia learning objects (LO) in the e-learning system. However, the problem is that the system is unable to allocate appropriate learning objects to the learners. Personalized learning environments make the system more adapt to the learner profile, thus improve their performance and quality of learning. The learning style (LS) of a learner is a prominent metric to understand the learner profile. This paper discusses an approach to personalize the content recommendation based on the learning style of the learner. A rule-based expert system is implemented to recommend the content to the learner, where the learner is modeled using a probabilistic learning style model, and the teaching aspects of learning objects are modeled using specific fields of IEEE Learning Object Metadata Standard. The rule-set defined in this paper is used to recommend the most relevant learning objects to the learners. Finally, the recommendations are cross-validated with the LO ranking from a set of 48 participants and found that 75% of recommendations are compatible with the learner choice.