Artificial Intelligence Approaches For Improved Adaptability in an Adaptive E-Learning Environment: a Review

Abstract

The concept of e-Learning, which has emerged with the rapid advancement in technology, is a crucial aspect in the field of education. The major issue with the traditional concept of e-learning is that it delivers information to all students in the same manner, irrespective of their individual learning requirements. Adaptive e-Learning Systems, which emphasise the significance of the differences in individual learning styles in modelling the ideal learning environment, attempts to bridge the gap between the student and the instructor that can be observed in a traditional e-learning environment by identifying and catering to individual learner requirements and capabilities. Artificial Intelligence techniques which have the ability to replicate the decisionmaking process of humans, are significant in the domain of adaptive e- Learning as they can be used to improve the adaptivity of the system. This paper assesses the Artificial techniques; Fuzzy Logic, Neural Networks, Bayesian Networks and Genetic Algorithms, emphasising their contribution towards the concept of adaptivity in the context of Adaptive e-learning. The study indicated an increase in the adaptation of Fuzzy Logic techniques, specifically Type 2 Fuzzy Logic Systems, and Bayesian Networks in the development of the Student Model in order to deal with the uncertainty of learning and student diagnosing processes. The application of Artificial Neural Networks to overcome issues in the existing Adaptive E-learning Systems, has also been identified through this review where the application of feature extraction via the Neural Network approach is an effective methodology to be used in the development of the Adaptation Model of an Adaptive E-Learning System to extract the most appropriate characteristics that can be used to identify learning styles of learners.