Modeling User Knowledge for Adaptive Learning in E-Learning Environment

Recently, data mining methods has widely been used in e-learning. In fact, various researchers attempt, by adopting data mining method, to understand the learners more and more and accordingly foster adaptive learning. The studies have employed the learners’ characteristics, performance, learning style and cognitive style. This study identifies the level of learners’ knowledge and develops models for them for customizing the material taught. To develop the suggested model, the researchers applied improved RBF neural network and used a three-step teaching approach to strengthen this network. This approach adopts the particle group optimization algorithm in the first step, the nearest neighboring K algorithm in the second step, and fastest reduction algorithm in the third step. Further parts of the article deal with the learners’ characteristics and four appropriate characteristics were developed for the purpose of predicting the class variable to determine the level of learners’ knowledge. In order to assess the suggested model, an online course of Excel was tested. The learners of this course were divided into three experimental groups. One of the groups employed the suggested model, and the other two groups of learners benefited from presentation of the mediocre and algorithm courses available in the research literature. Test results suggest academic achievement and satisfaction of the learners paced in the group relating to the suggested model.