Thesis (M.Sc.) - Cairo University - Faculty of Computers and Artificial Intelligence - Department of Computer Science

Adaptive learning is one of the most widely used data driven approach to teaching and it received an increasing attention over the last decade. It aims to meet the student{u2019}s characteristics by tailoring learning courses materials and assessment methods. In order to determine the student{u2019}s characteristics, it is necessary to detect their learning styles according to VAK (Visual, Auditory or Kinesthetic) learning style. Adaptive learning includes the adaptation of all learning components. Assessment is one of the most important learning components that could enhance the learning outcomes. Implementing alternative types of assessment to meets students{u2019} characteristics and abilities lies under the umbrella of adaptive learning. In this research, an integrated model that utilizes both semantic technology and machine learning clustering methods is developed in order to cluster students and detect their learning styles then recommend suitable assessment method(s) accordingly. In order to measure the effectiveness of the proposed model, a set of experiments were conducted on real dataset (Open University Learning Analytics Dataset). Experiments showed that the proposed model is able to cluster students according to their different learning activities with an accuracy that exceeds 95% and predict their relative assessment method(s) with an average accuracy equals to 93%

Issued also as CD