Automated negotiation might improve negotiation outcomes in e-commerce and allow the coordination of agents in autonomous systems. However, as operative systems do not yet exist, research relies on simulation to evaluate potential system configurations in realistically complex scenarios. This book reviews the latest literature on the simulation of automated negotiation and addresses identified deficiencies by offering novel strategies for software agents and alternative interaction protocols. The method of computer simulation is discussed, with a focus on its specific applications in the domain of negotiations. The results of computer simulations of the proposed automated negotiation systems are compared across systems and to the outcomes of negotiation experiments with human agents. Analyses along various outcome dimensions, such as efficiency, fairness and individual utility, provide insights into general design trade-offs as well as design requirements for superior outcomes in automated negotiations.
Michael Filzmoser received his MSc in international business administration and his PhD in Management from the University of Vienna. He is currently university assistant at the institute of management science of the Vienna University of Technology. His teaching and research interests lie in simulation, negotiation and organization theory.