Indonesia Conference Directory

In multi-attribute automated negotiation, maximizing negotiation outcome can be regarded as a multi-objective optimization problem. The objectives are to maximize each negotiator gain expressed as their preferences. A negotiation outcome is evaluated based on the outcome distance to the Pareto front of all the possible negotiation solutions. Pareto front is the collection of solutions that can no longer be improved in one objective without compromising another objective. Therefore Pareto front plays an important role in examining negotiation outcomes. In cases where possible negotiation solutions are too large, it is computationally inefficient to find this front exhaustively by comparing all solutions. To this end, evolutionary algorithm has been used to approximate Pareto front. In this paper we aimed to obtain Pareto front using an evolutionary algorithm called non-dominated sorting genetic algorithm II (NSGA II). Results using two different negotiation settings show that Pareto front obtained with NSGA II are very close to the true Pareto front.