You are here

A CP Approach to the Balanced Academic Curriculum Problem

TitleA CP Approach to the Balanced Academic Curriculum Problem
Publication TypeConference Proceedings
Year of Conference2007
AuthorsMonette, Jean-Noël, Schaus Pierre, Zampelli Stéphane, Deville Yves, and Dupont Pierre
Conference NameThe Seventh International Workshop on Symmetry and Constraint Satisfaction Problems (Symcon'07)
Date Published23/09/2007

The Balanced Academic Curriculum Problem (BACP) has received little attention in Constraint Programming. Only a few articles deals with this problem with experimental results on the three small instances publicly available in CSPLIB. The present article describes an approach to efficiently solve this challenging problem. Optimal solutions are produced on a variety of randomly generated instances which generalize the CSPLIB test cases. This work describes four contributions to the resolution of this problem: a new branching heuristic, the use of dominance relations, experiments on several balance criteria and several search strategies among which an hybridization of Constraint Programming and Local Search.

Full text: