New Constraint Programming Approaches for the Computation of Leximin-Optimal Solutions in Constraint Networks

Sylvain Bouveret, Michel Lemaître

Abstract: We study the problem of computing a leximin-optimal solution of a constraint network. This problem is highly motivated by fairness and efficiency requirements in many real-world applications implying human agents. We compare several generic algorithms which solve this problem in a constraint programming framework. The first one is entirely original, and the other ones are partially based on existing works adapted to fit with this problem.