Modeling Personalized Learning Styles in a Web-Based Learning System

Abstract

An innovative learning mechanism for identifying learners’ learning styles to improve adaptive learning is proposed. Hypermedia-learning tools are highly interactive to learners in web-based environments that have become increasingly popular in the field of education. However, these learning tools are frequently inadequate for individualize learning because accessing adaptive learning content is required for learners to achieve objectives. For predicating adaptive learning, a neuron-fuzzy inference approach is used to model the diagnosis of learning styles. Then, according to the diagnosis results, a recommendation model is constructed to help learners obtain adaptive digital content. The proposed approach has the capability of tracking learning activities on-line to correspond with learning styles. The results show that the identified model successfully classified 102 learners into groups based on learning style. The implemented learning mechanism produced a clear learning guide for learning activities, which can help an advanced learning system retrieve a well-structure learning unit.