A rational multiparty information exchange model using extensive games

Abstract

It is well known that excessive intelligence is mandatory in military procedures, in companies to know about competitors, as well as in markets to decide where to invest in. However, this information is not always fully reliable since it might be given by others with conflicting interests. Consequently, there is no guarantee that the supplied information has not been maliciously crafted to produce negative outputs to the receiving entity. For a formal treatment, a way to represent this problem is via Multi-Party Information Exchange (MPIE) model. Applying extensive games, we propose a MPIE when the information is relayed in non-secret channels. This is a refinement of a previous MPIE by rejecting solutions arisen from empty threats, through finding perfect equilibria in sub-games. The resolution to find such perfect equilibria is based on executing an algorithm that solves Mixed Integer Linear Programming (MILP) models. The case studies prove that empty threats are deleted and that a large number of players can be efficiently solved using the proposed algorithm.