Agent based social simulation

Abstract

One of the applications of agent technology is the simulation of social systems. Agent-based simulation tools have been developed with this purpose. However, although the agent paradigm should facilitate the implementation of models for social simulation, because of its intentional and social features, the definition of these models and the use of these tools requires a deep knowledge of programming languages, something that is not very common in social scientists. This makes necessary the availability of new modelling tools, which shold be easier to adapt to the needs of sociologists, with basic programming skills. Furthermore, social simulation tools should support higher degrees of scalability, concerning both the edition of the simulation models and their execution. In order to cope with these requirements we propose the development of graphical simulation environments, which should be application domain oriented. Their implementation can be done using an open and flexible agent development framework, INGENIAS, which provides concepts, methods and tools for the development of multi-agent systems and code generation on different platforms. © AEPIA.