AN IMPROVED TIMETABLE SCHEDULING USING CONSTRUCTIVE ALGORITHM

Timetable scheduling remains a critical yet challenging task in educational institutions, where conflicting constraints such as resource availability, faculty preferences, and student requirements often lead to inefficiencies, manual rework, and suboptimal allocations. Traditional approaches, including manual methods and basic constructive techniques, frequently result in fragmented schedules that fail to optimize utilization or satisfy all stakeholders. This project proposes an improved timetable scheduling framework leveraging a constructive algorithm as the core scheduling technique to address these limitations. The constructive algorithm employs a step-by-step greedy construction mechanism, iteratively assigning courses, instructors, and venues while dynamically evaluating partial solutions against a multi-objective fitness function. Key enhancements include integration of constraint propagation for early conflict detection, adaptive prioritization of hard and soft constraints (e.g., no overlapping classes for instructors), and a post-construction refinement phase using local search to minimize violations. Preliminary evaluations on benchmark datasets from university scheduling scenarios demonstrate a 25–35% reduction in scheduling conflicts compared to baseline genetic algorithm methods, achieving over 95% constraint satisfaction within reasonable computational time (under 5 minutes for 500-course instances). This approach not only enhances scalability for large-scale environments but also provides a flexible, open-source tool for automated timetable generation. Future extensions could incorporate machine learning for predictive constraint forecasting, paving the way for real-time adaptive scheduling in dynamic educational settings.