On the Problem of Recognizing and Learning Observable Social Exchange Strategies in Open Societies

摘要

Regulation of social exchanges refers to controlling social exchanges between agents so that the balance of exchange values involved in the exchanges are continuously kept - as far as possible - near to equilibrium. Previous work modeled the social exchange regulation problem as a POMDP, and defined the policy To BDI plans algorithm to extract BDI plans from POMDP models, so that the derived BDI plans can be applied to keep in equilibrium social exchanges performed by BDI agents. The aim of this paper is to extend that BDI-POMDP agent model for the self-regulation of social exchanges with a HMM-based module for recognizing and learning partner agents' social exchange strategies, thus extending its applicability to open societies, where new partner agents can freely appear at any time. For the recognition problem, the BDI-POMDP-HMM agent proceeds by analyzing the patterns of refusals for exchange proposals that are present in a partner agent's behavior. For the learning problem, it learns HMM to capture probabilistic state transition and observation functions that model the social exchange strategy of the partner agent. The agent then transforms the HMM's transition and observation functions into POMDP's action-based state transition and observation functions, obtaining a POMDP model of the partner's previously unknown social exchange strategy, and deriving corresponding exchange regulation plans through policy To BDI plans. The paper also presents a discussion of the results of some simulations.