An automated school timetabling system using hybrid intelligent techniques

Abstract

School timetabling is typically hard to resolve and the problem gets harder when teaching resources are scarce. The need for supporting stringent teaching requirements such as co-teaching and split-class teaching complicates the problem even more. In this paper, we describe an automated school timetabling system called @PT that reasons about sophisticated timetabling requirements. @PT makes use of a number of intelligent techniques including constraint technology, heuristics, local search operators, and tabu-list like data structure. It also changes its search behavior dynamically at run-time. Experimental results show that @PT is robust and manages to solve real problem instances effectively within minutes.