A Computational Toolset for Socio-Cultural Data Exploration, Model Refinement, and Model Validation

Neal Reilly, S., Pfeffer, A., Barnett, J., Chamberlain, J., and Casstevens, R.

Proceedings of Human Social Culture Behavior (HSCB) Focus 2011, Chantilly, VA (February 2011)

Accurate computational models of social and cultural behavior have the potential to dramatically improve strategic planning by providing decision makers with insights into the behavior of adversaries and contested populations. The Human Social Cultural Behavior (HSCB) modeling community, which is working to create such models, has focused most of its effort on developing tools to support the computational implementation of social science models and relatively little effort on developing tools to help social scientists ensure that the underlying social science models are valid or to understand the limitations of their effective use. We will describe an ongoing effort to develop a suite of tools called the Model Analyst’s Toolkit (MAT). MAT is aimed at social scientists who are developing and refining their models, who need to effectively and deeply validate their models against available data, who need to understand and communicate the limitations of the validation process due to a host of factors, and who need to effectively hand off their model for implementation by computational scientists and for use by strategic decision makers. MAT comprises graphical tools to explore data, define causal structures in models, and engage in a collaborative human-machine validation and analysis process.

For More Information

To learn more or request a copy of a paper (if available), contact S. Neal Reilly.

(Please include your name, address, organization, and the paper reference. Requests without this information will not be honored.)