Improving Students Academic Performance Using Hybrid Recommendation Techniques

Abstract

Development of students' performance is significant in educational environments because it plays an essential role in making the best quality graduates and post-graduates who will become great leaders in the future and sources of the workforce for the country. A recommendation system is an intelligent system that proposes different suggestions to students, based on the previous actions from other students who faces the same environment, such as academic performance. One of the major problems today the high rate of students failure is a worry for many universities. This study proposed recommendation system to identify weak academic students as soon as possible to help them in a suitable time, encourage students to study hard when they know that they are at risk and to plan their workload carefully. The study is applied the hybrid recommendation system that is one approach for recommendation system. This approach is executed by used each of clustering algorithms and association rules algorithms on the nature of data which have been collected from the University of Kordofan, Faculty of Computer Studies and Statistics. The clustering algorithms results were evaluated regarding to high accuracy for each cluster and then applied Association rules algorithms in particular. The obtained results are generated strong rules that appear which courses are effectiveness posative or nagative on accumulative GPA.