Automatic mechanism design for active service environment

In oriented service computing area, the applications of theory and technology of multi-agents can make it possible to change computing model from a centralized controlling type to a distributed cooperation form. In the meantime, we have to take some steps to constrain and lead global behaviors of multi-agents so as to ensure that an effective consultation alliance will bring out smoothly, for agent with self-interest feature in the traditional multi-agents theory. This paper presents an approach for proposing solutions to specific requirement based on Automated Mechanism Design(AMD) techniques. The main contribution is to describe how to built up the computational settings in mechanism designing process for active service environment, such as agency group, type sets, special outcomes for requirement, utility functions and an objective goal. After that, a simulation experiment is put forward to verify the rationality of mechanism designed in this paper as an ending.