Course Recommendation System: Collaborative Filtering, Machine Learning and Topic Modelling

Abstract

Course Recommendation System is required in the education institution to recommend the course to the student based on interest and preferences. Various existing methods were applied for the course recommendation system to improve recommendation reliability. This study reviews the existing methods in the course recommendation system and its performance in the course recommendation. Most of the methods were applied the personalized information like student score, interest and learning style in the recommendation model. Machine learning methods were commonly applied for the classification of the data to improve the recommendation performance. Topic modeling was applied in some research to analyze the student interest in the course. Collaborative filtering methods analyze the similarity between the students and course to provide an effective recommendation. The Random forest and decision tree models applied in the course recommendation involve an overfitting problem and unstable performance. The review shows that machine learning-based and collaborative-based models have higher performance in recommendation systems. Still, cold start, and overfitting problem in the course recommendation system needs to be overcome to provide effective performance.