A Survey of Genetic Algorithms for the University Timetabling Problem Manar Hosny and Shameem Fatima College of Computer and Information Sciences, King Saud University, Riyadh, Saudi Arabia mifawzi@ksu.edu.sa , sfatima@ksu.edu.sa Keywords: Timetabling, Scheduling, Combinatorial Optimization, Genetic Algorithms, Heuristics, Meta-heuristics. Abstract. The university timetabling problem is a scheduling problem that many educational institutions need to solve in order to plan their courses and/or exams by allocating these events to specific timeslots, rooms, lecturers…etc. There are also a number of hard and soft constraints that must be observed while solving this problem, which makes the solution algorithm a challenge for researchers. Genetic Algorithms (GAs) is a popular meta-heuristic technique that has been successfully applied to many hard combinatorial optimization problems, among which are timetabling and scheduling problems. In this paper we describe some GA techniques that have recently been applied to different variants of the university timetabling problem producing promising results. We briefly describe the overall technique, focusing on the chromosome representation and the crossover and mutation operators. A summary of the described algorithms and their most distinguishing features is then presented at the end of this paper. Introduction The timetabling problem is an important practical problem that is frequently encountered in educational institutions, such as schools and universities. The timetabling problem has received special attention from the scientific community in the last few decades. This is mainly due to the fact that manual generation of timetables is very time consuming and the resulting timetables are usually inefficient and may be costly in terms of money and resources. The interest in timetabling algorithms resulted in the creation of the PATAT series of conferences (Practice and Theory of Automated Timetabling), which sponsors the International Timetabling Competition (ITC) [1]. The aim of this competition is to encourage research in the university timetabling domain and bridge the gap between theory and practice, for a better utilization of research techniques in real-world applications. Scheduling and timetabling problems have many forms like constructing timetables for exams and courses in educational institutions, scheduling of employees’ shifts and working hours, scheduling of sports or business events … etc. In this paper we concentrate on one class of timetabling problems, which is University Timetabling. In this problem type, it is required to allocate a number of events (such as exams or courses or student groups) to a number of available classrooms and a number of timeslots, while adhering to a set of pre-specified problem constraints. In general, the constraints are classified into two types: hard constraints and soft constraints [2]. Hard constraints must be strictly enforced in any solution, i.e., they cannot be violated. For example, the condition that a student cannot attend two classes (or two exams) at the same time is a hard constraint. Solutions that violate hard constraints are infeasible solutions. On the other hand, soft constraints can be accepted in a solution, but are not preferred.