journal-of-information-and-optimization-sciences Journal of Information and Optimization Sciences 2169-0103 0252-2667 Taru Publications 10.47974/JIOS-2154 A machine learning approach for personalized course recommendation systems for learners Thomas Anu Department of Computer Engineering, Lok Jagruti Kendra University, Ahmedabad, Gujarat, 382210, India Pandi Gayatri Department of Computer Science and Engineering, Sindhu Bhavan Road, New L J Institute of Engineering and Technology, Ahmedabad, Gujarat, 380059, India 1 11 30 06 2026

This paper presents a novel approach to enhancing collaborative filtering for course recommendations by integrating Hybrid Fruit-fly Optimized Density-Based K-Means Clustering (HFO-KMC) with Atten-BERT-RU and a Coordinate Attention Module (CAM). The methodology improves user grouping and feature representation, resulting in more accurate and personalized recommendations. HFO-KMC optimizes clustering by reducing errors, while Atten-BERT-RU with CAM enhances feature extraction from course reviews. The system was validated using Coursera course reviews, showing notable improvements in recommendation accuracy and diversity over traditional models. The results highlight the effectiveness of combining optimized clustering with deep learning techniques, making this approach a significant advancement in recommendation systems.

 Recommendation systems Machine learning Collaborative filtering Content-based filtering Deep learning Fruit-fly optimization K-means clustering Personalized learning Educational data mining Coursera BERT Attention mechanism Hybrid models Course recommendation User modeling access open retracted no