Evidence-Driven Multi-Objective Optimisation for MOOC Update Planning: A Case Study in Cyber Security Education

Massive Open Online Courses (MOOCs) have expanded access to education, but many courses continue to face problems such as learner dropout and uneven engagement. One reason is that course updates are not always guided by learner feedback. This study examines how learner feedback can support systematic MOOC updates. An optimisation framework is proposed to help instructors decide which course elements to revise under limited time and effort. The framework considers three aspects: learner value, instructor effort, and diversity of updates. Survey responses were collected from 437 learners enrolled in the July 2025 offering of the Introduction to Cyber Security MOOC on SWAYAM. Based on this feedback, 26 candidate updates were identified and described using learner value, effort, and category. The Non-Dominated Sorting Genetic Algorithm II (NSGA-II) was applied under five effort budgets (8–16 units) to generate alternative update plans. The results show that learner value increases rapidly at lower effort levels but slows beyond 12 effort units. Diversity also improves initially and then stabilises. Comparison with a greedy method and an instructor-designed plan shows that the optimisation approach produces more balanced update decisions. The study shows that structured optimisation can support clear and evidence-based MOOC revision.