Exam time-table scheduling

Abstract

Exam timetable scheduling is a complex problem which often has to be dealt with at Universities. This project investigates this problem for the Faculty of ICT in University of Malta (UoM). Currently, examination officers start this process several weeks before examination period. The work involves dedicating approximately 2 working days to produce an initial draft, and performing changes upon student and/ or staff complaints, until a final timetable is prepared, possibly weeks later. This dissertation demonstrates how, using Genetic Algorithms, this process can be complete in just about 1 minute, while also improving the quality of timetables from those generated by hand. Apart from generating a timetable which is conflict free and significantly minimizes near exam occurrences for students during this stressful period, an intuitive user interface, with a calendar design, is also provided to aid examination officers through the whole process of producing a final timetable. This includes making the whole system interactive and providing useful information to accommodate for late changes, optional units from other faculties, and special requests which are especially common in real-word problem applications.