A Simulation Environment to Support Simulation Studies on Global Public Goods

Abstract

This paper presents an agent-based simulation (ABS) environment, GPGSiM, which aims at supporting the simulation-based studies on international politics and economics. Its key feature is the conceptual model that consists of such classes as Model, Agent, Dyad, Space and Geometry. It is defined based on the analysis of actual research processes in those domains, by cyclically performing the design of the system and the study of individual problems in the domain, which we call bottom-up approach to design an ABS environment. GPGSiM provides a class library, simulator, and an Eclipse plug-in. The class library implements the conceptual model and enables the users to build their own models effectively. The simulator provides support for running simulations as well as visualization tools for analysis, and the Eclipse plug-in enables smooth coordination of the subsystems. Various models in the domain are implemented, including well-known models, such as Spatial Prisoners’ Dilemma, and original ones. In order to evaluate the efficiency of model development in GPGSiM, we compared the size of codes of several models, and it is shown that our approach successfully frees the users from writing programs that are not essential to their own models. We also propose a system, GPGCloud, which helps users to share their simulation models with others, to cope with the programming related cost. It also provides an easy access to a large computation environment that allows them to simultaneously perform a number of simulation runs based on their models, which reduces the cost on managing large computers. It supports researchers in simulation-based study of international politics and economics in combination with programming supports provided by GPGSiM.