Ontology-Driven Project Recommendation for Intelligent Recommender System
Asma Hadyaoui, Lilia Cheniti‐Belcadhi · Interactive Learning Environments · 2025
The growing intersection of educational technology and personalized learning has led to the development of sophisticated Recommender Systems (RSs) that enhance learner engagement. This article introduces PEARL (Project Engagement and Affinity Recommender with Learning), a novel ontology-driven RS designed for Project-Based Collaborative Learning (PBCL) environments. PEARL integrates collaborative filtering, decision algorithms, and ontology-based personalization to generate context-aware project recommendations that align with learners’ competencies and pedagogical goals. At the core of PEARL is an intelligent recommendation engine that employs feedback loops to iteratively refine predictions by analyzing learner interactions, project characteristics, and group dynamics. Empirical validation conducted in a Python programming course demonstrated the framework’s effectiveness, achieving precision and recall rates of 82% and 78%, respectively. These metrics highlight PEARL’s ability to deliver relevant and engaging project suggestions while maintaining a balance between individual personalization and collaborative team formation. PEARL has shown significant potential to improve learner engagement and project completion rates, establishing it as a valuable advancement in educational technology. These findings contribute to the development of adaptive and intelligent RS tailored for PBCL environments.