Intelligent agents for negotiations in market games. I. Model

摘要

In competitive markets, human decision-makers need assistance tondetermine potential coalitions that they can be part of. Note that suchncoalitions are legal provided market power is not monopolized. Wenpresent an approach for designing specialized intelligent softwarenagents that perform negotiations on behalf of their human counterparts,nand then suggest market strategies that the human can adopt. Ournnegotiation protocol is derived from cooperative game theory; however,nit differs from the game theory literature in requiring virtually nontrustworthy information exchange between the potential coalitionnpartners. In this paper, we present the negotiation model