Can analytic models, informed by social scientific theories using computational engineering approaches, offer effective forecasting of violent behavior? This chapter discusses a new data set which codes the structure and behavior of ethno-political organizations and the use of a new approach for forecasting political behavior drawn from computer engineering. In the chapter, we build a forecasting model and then test the model against existing data as well as a predictive analysis for the year 2009 (the analysis was done in 2008 and data for 2005–2009 has not yet been collected for this data set). The data used was drawn from the Minorities at Risk Organizational Behavior (MAROB) data set. MAROB was created through collaboration between the National Consortium for the Study of Terrorism and Responses to Terrorism and the Minorities at Risk (MAR) Project. This data focuses on ethno-political organizations in the Middle East to test factors that make it more or less likely that an organization will choose to use violence. While the variables on which data was collected were informed by theories of contentious politics, this chapter focuses primarily on the data itself and the forecasting approach that we used and less on the social science theoretical models as such. Analytically we use multiple approaches for data massaging, classification and forecasting to achieve high classification accuracies (measured in terms of overall accuracy, recall, precision, false positives, and F-measure). We also strive for parsimony in the number of variables we use to make our forecasting predictions.
Asal, Victor, and Kihoon Choi, Krishna Pattipati, Jonathan Wilkenfeld. 2013. "Forecasting the Use of Violence by Ethno–Political Organizations: Middle Eastern Minorities and the Choice of Violence." In Handbook of Computational Approaches to Counterterrorism, ed. V. S. Subrahmanian. New York, NY:Springer, 221-225.