Deep Learning-Driven Personalized Course Recommendations in MOOC Platforms Using Domain-Specific Strategies

Course recommendations on online platforms are full of uncertainty and ambiguities due to the involvement of several uncertain factors. The primary goal of this study is to investigate an intelligent fuzzy decision algorithm that can assess learners and enhance personalized instruction within Massive Open Online Courses (MOOC) platforms. Several frameworks model uncertain situations, but the T-spherical fuzzy set (TSFS) can deal with uncertainties in data within diverse contexts. The algorithm discussed in this study can assess multiple learning performance metrics to enhance learners’ abilities, recognize strengths, and identify areas for improvement. The instructors can quickly evaluate and select training modules customized to particular requirements using the designed algorithm. The algorithm assesses learner performance using a quantitative evaluation that includes Learning Proficiency Level (LPL), Engagement Metrics (EM), Conceptual Understanding (CU), Consistency in Learning (CL), Response Time (RT), Analytical and Critical Thinking (ACT), Adaptability to Course Content (AC), and Collaboration Skills (CS). The results inform the customization of training programs for learners. This research examines strategies to decrease dropout rates in MOOCs by tailoring learning experiences to align with learners’ motivational factors. The process discussed in this study proposes a novel methodology for profile construction in a MOOC recommendation system, incorporating external data to delineate a learner’s interests.