Agent-based Modeling of Culture's Consequences for Trade

Cultural differences are known to have their effects on trade. The subject of this thesis is a computer simulation of the role of trust and deceit in trade across cultures. In the computer simulation the traders’ rolls are realized by software agents. Such an agent is a computer program which simulates the behavior of a human being. In a multi-agent simulation a group of software agents is acting and interacting simultaneously. Important functions of agents in the present simulation are to approach new potential trade partners, to negotiate about a transaction and to exchange proposals, and, when the negotiation has ended successfully, to exchange products, and to decide and request a trace to be performed. The agents’ decision mechanisms are implemented according to models and data available from scientific literature. To model the influence of culture on the decision making, an expert systems approach is taken, using the Synthetic Cultures according to Hofstede en Pedersen. The approach has proved feasible to develop a valid model of cultural differentiation in software agents. This is relevant for the development of instruments for cross-cultural research, educational and training applications to make people aware of cultural differences, and affective human-computer interfaces in a globalizing world.

D. Verwaart

8-6-2011 12:30:00