Abstract

This paper describes our work on applying novel techniques based on propositional satisfiability (SAT) solvers and optimizers to the Curriculum-based Course Timetabling problem. Out of 32 standard benchmark instances derived from the Second International Timetabling Competition held in 2007, our techniques yield the best known solutions for 21 of them (19 of them being optimal), improving the previously best known solutions for 9. In addition, we obtain 18 new lower bounds for this benchmark set by applying a new full (Weighted) Partial MaxSAT approach of the Curriculum-based Course Timetabling problem.