Scheduling social tournaments locally

Issue title: Constraint Programming for Planning and Scheduling

Article type: Research Article

Authors: Dotú, Iván | Van Hentenryck, Pascal

Affiliations: Departamento De Ingeniería Informática, Universidad Autónoma de Madrid, Spain | Brown University, Box 1910, Providence, RI 02912, USA

Abstract: Tournament scheduling, such as the social golfer problem, has attracted significant attention in recent years because of their highly symmetrical and combinatorial nature. This paper presents an effective local search algorithm for a variety of tournament scheduling problems, including social golfer problems, debating tournaments, judge assignments and very social golfers. The algorithm finds high-quality solutions on all problems, including new solutions to open problems. Interestingly, the algorithm does not incorporate any symmetry-breaking schemes and is conceptually simple when compared to advanced constraint-programming solutions.

Keywords: Scheduling, social golfer, local search, constraint programming

Journal: AI Communications, vol. 20, no. 3, pp. 151-162, 2007