Interactive visualization of agent societies

Abstract

Visualization of the dynamic behavior in complex systems is an efficient way to facilitate analysis and decision making. Complexity itself entails several difficulties to find an adequate visual representation of the relevant features of the system's dynamics. For this reason, in this work we explore the elaboration of interactive interfaces for the visualization of the behavior of complex systems. In particular, we pursue the representation of the dynamics of a society of intensional agents. These agents exhibit an individually simple behavior, but together experiment a collective complex behavior. We established an experimentation protocol aimed to verify an hypothesis about the long term evolution of the system. We show how our visualization tool is helpful in understanding and interpreting these results.