Implementing Kempe chain neighborhood structure in harmony search for solving curriculum based course timetabling

An essential aspect that contributes to the success of meta-heuristic algorithm over a curriculum-based course timetabling problem is determined by the neighborhood structure used. The basic neighborhood structures such as move and swap between lectures has no method for escaping from local minima or optimum that restricts the improvement of current solutions. The aim of this paper is to implement Kempe chain neighborhood structure together with the other neighborhood structures in harmony search algorithm for solving curriculum-based course timetabling. The result shows significant improvements in the solution quality.