Integrating an e-learning model using IRT, Felder-Silverman and Neural Network approach

Abstract

Personalized learning seek to provide each individual learner with the right and sufficient content they need according to learners level of knowledge, behavior and profile. One of the most important factors for improving the personalization methods of e-learning system is to apply adaptive properties. The aim of adaptive personalized e-Learning system is to offer the most appropriate learning materials to learners by taking into account their background and profiles. However, most of the systems focused on users' learning behaviors, interests and habits to provide personalized e-Learning services while ignoring course difficulty, users profile and user's ability. Recent researchers focus on fuzzy implementation of item response theory to measure learner's ability and course difficulty. This paper introduces an improved model by using a personal e-Learning by integrating Item Response Theory and Felder-Silverman's learning style theory as an attempt to obtain personal knowledge, background and learning style. These input will be verified and classified by an Artificial Neural Network as machine learning to model their behavior as whole. This technique will be able to estimate the ability of students towards improving the level of understanding to moderate until weak students in programming classes. Therefore, there will be suggestions for course materials suitable for students and course material difficulty can be adjusted automatically. It is hoped that this study will contribute towards higher education institution for an adaptive e-Learning rather than content-focus e-Learning.