Agent-Based Modeling and Simulation of Collaborative Social Networks

Gregory R. Madey, Yongqin Gao, Vincent W. Freeh, Renée Tynan, Chris Hoffman

Journal of the Association for Information Systems · 2003 · 16 citations · 20 references

Full text

Open access

Concepts

Abstract

We describe a research framework for studying social systems.The framework uses agent-based modeling and simulation as key components in the process of discovery and understanding.A collaborative social network composed of open source software (OSS) developers and projects is studied and used to demonstrate the research framework.By continuously collecting developer and project data for over two years from SourceForge, we are able to infer and model the structural and the dynamic mechanisms that govern the topology and evolution of this social network.We describe the use of these empirically derived agent-based models of the SourceForge OSS developer network to specify simulations implemented using Java/Swarm.Several network models and simulations of the evolution of SourceForge, and the verification and validation processes of the framework are described.The nature of social network processes hidden from view that could plausibly generate the observed system properties can be discover through an iterative modeling, simulation, and validation and verification process.Such a process, dynamic fitness based on project life cycle, was discovered.

References

20