AI-Driven Hybrid Recommender Framework for Personalized Course Selection in Learning Platforms

The rapid expansion of digital learning platforms has resulted in an overwhelming number of online courses, making it increasingly difficult for learners to identify offerings that align with their academic goals, skills, and interests. This study addresses this challenge by proposing a machine learning–based hybrid course recommendation framework that integrates content-based filtering and collaborative filtering to enhance personalized learning support. The proposed approach combines statistical and semantic text representations using Term Frequency–Inverse Document Frequency (TF–IDF) and Sentence-BERT (SBERT) embeddings, together with regression-based predictive modeling for learner preference estimation. Content-based filtering captures semantic similarity between learner profiles and course descriptions, while collaborative filtering predicts course ratings based on historical learner–course interactions, enabling a balanced and context-aware recommendation process. The framework is evaluated using two independent datasets: a large-scale public Coursera dataset for benchmarking and a real-time institutional dataset collected from engineering colleges in Maharashtra for practical validation. Experimental results indicate that models leveraging SBERT embeddings consistently outperform TF–IDF-based representations across multiple regression models, with the SBERT-Gradient Boosting Regressor combination achieving the most reliable performance in terms of Mean Absolute Error (MAE), Mean Squared Error (MSE), and R². These findings suggest that integrating semantic representations with predictive modeling improves recommendation accuracy and robustness across heterogeneous educational contexts. While a graphical user interface is presented to demonstrate real-world applicability, the primary contribution of this work lies in the methodological framework and its empirical validation. Overall, the proposed hybrid recommender framework offers a scalable and interpretable solution for personalized course selection and supports flexible, interdisciplinary learning objectives aligned with India’s National Education Policy (NEP) 2020.