Danilo Saft

By Danilo Saft and Volker Nissen

The field of complexity science often employs multiagent simulations to investigate complex and emergent behavior. Authors in complexity science have suggested that the discussion of complex systems could benefit from a more systematic approach and a more compact mathematical way to describe the behavior of such systems in addition to the common observations and interpretations taking place today. Regarding quantitative measures to capture emergent phenomena, several approaches have been published, but have not yet been put to wide systematic use in the research community. One reason for this could be the manual effort required to investigate multiagent systems in a quantitatively accurate form. Toward this end, there has so far been a lack of appropriate and easy-to-use IT-based tools. To eliminate this deficiency, we present a software library, which enables researchers to integrate emergence measurements into experiments with multiagent modeling tools such as Repast and NetLogo. The major benefit for researchers is that this toolbox enables them to make comparable, quantitatively well-grounded statements about the emergent behavior of the model at hand. The toolbox therefore provides researchers with a standardized artifact that can be employed in a systematic methodological approach to the analysis of multiagent systems.

Part of the book: Multi-agent Systems