Scheduling meetings with distributed local consistency reinforcement

Abstract

Meeting scheduling (MS) is an important real-world problem. Solving this problem consists in scheduling all the meetings while satisfying all the constraints. However, human nature often has conflicting preferences. The majority of works, dealing with MS problem, allowed the relaxation of the preferences in order to reach an agreement between all the participants, but this is not always possible. To overcome this difficulty, the main contribution of our work consist in trying to satisfy as much as possible users' preferences while taking into consideration their availabilities, and this through a new approach based on the distributed reinforcement of arc consistency (DRAC) model. The new approach was implemented and the experimental results show that our approach is scalable and worthwhile to handle especially strong constraints.