An approach for e-learning data analytics using SOM clustering

Abstract

Within the field of e-learning, the accessibility of large amount of learning materials in web-based education systems often burden students to get the learning materials that they preferred. Recent study in e-learning has intensified the need for adapting students' behaviour and knowledge level in presenting learning materials to the students. However, the heterogeneous of students' behaviour, the diversity of learning materials and the complexity of hyperlinks for navigations become the limitation attributes of an adaptive learning environment system. Hence, there is a need to develop techniques to analyse the students' data in order to understand the student's behaviour in order to organize and maintain the learning materials in the domain model repository in the system. In this research, we propose a framework for adaptive learning environment by using self-organizing map (SOM) clustering approach for learner model data analytics and structuring the domain model content in order to present the learning content that suits with the student's need and to achieve higher performances in learning. This research contributes to big data analytics in e-learning which currently is the new focus among educators as the learning data grows tremendously every day