On Adapting a War-Gaming Discrete Event Simulator with Big Data and Geospatial Modeling Toward a Predictive Model Ecosystem for Interpersonal Violence

摘要

The United States leads industrialized countries in rates of interpersonal violence with homicide being the second leading cause of death for people aged 15 to 24 years. In 2010, more than 4,800 youths (ages 10 to 24) received emergency treatment at hospitals due to injuries caused by physical assaults. This problem has taken epidemic proportions with 33% of high school students reporting physical altercations within the last year, 20% reporting being bullied on school grounds, 16% reporting electronic bullying, and 5% declaring that they had taken a weapon to school within the last 30 days prior to completing a survey conducted by the Centers for Disease Control in 2012. This paper presents an approach to adapting a war-gaming discrete event simulator with big data and geo-spatial modeling towards construction of a predictive model ecosystem for interpersonal violence. The ecosystem will be designed and tested using United States data on interpersonal violence collected over the past 20 years. Spatio-temporal data on interpersonal violence will be collected across the entire United States and stored in a Big Data management and analytics facility that will provide the basis for mapping the patterns of historical and current interpersonal violence. The facility will contain both analytical and simulation tools that collectively allow the researcher to input a strategy and observe predicted future states. The adapted discrete event simulation facility is envisioned to use a predictor-corrector method which will make the ecosystem a self-improving model for interpersonal violence prediction.