Course Recommendation System for Students Using K-Means and Association Rule Mining

Abstract

Students have difficulty choosing the most suitable courses during their undergraduate studies. In the academic world, the institutions offer students various courses to study. They have several options to choose from many courses based on their future career planning, interest, and advice from peers, seniors, teachers, etc. Hence, inappropriate course selection leads to innumerable difficulties, poor performance and dissatisfaction. Thus, it's essential to propose a course recommendation system that helps undergraduate students in the course selection process. This paper presents a machine learning approach to recommend relevant courses to students based on popular courses. The K-Means clustering method has been used to find students' most and least demand courses. Then, the FP-Growth algorithm generates the rules to recommend suitable courses for a specific student. A real-world dataset has been used, which consists of undergraduate students' academic records. The proposed method is evaluated by applying the dataset that would perform relatively better.