Personalized Course Recommender System Based on Multiple Approaches: A Comparative Analysis

Nowadays, online learning platforms offer an excellent opportunity for learners to access various courses from various fields, breaking down geographical barriers and promoting lifelong learning. However, it can be overwhelming for students to select the appropriate course among the numerous options available; it is also exhausting for a learner to find similar courses developing the same characteristics and skills as an already taken course. Thus, recommendation systems are crucial in helping online learners filter suitable content and make the right decision during course selection, which positively impacts student engagement and the overall learning experience. However, many widely used e-learning platforms have yet to incorporate recommendation engines. Additionally, educational platforms do not benefit from a sophisticated and accurate recommendation system like those found on streaming services, social media platforms, and online marketplaces. This paper presents a personalized course recommender system based on multiple recommendation approaches, including collaborative filtering, content-based filtering, popularity-based models, and hybrid models. The system is designed to aid online learners in selecting courses that match their interests with personalized content. It offers solutions for new users by suggesting popular courses and other users’ preferences. To ensure precision and evaluate the efficiency of each model, we evaluated their performance using a range of metrics, such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), Precision, and Recall, in a comparative context.