Big data might predict violent outbreaks Posted on : Mar 26 - 2015

The military could get a new method for anticipating violent outbreaks in tense areas such as Afghanistan, one that uses data and analytics rather trying to predict the future based on the past.

According to a report at SciDev.net, political scientist Jason Lyall of Yale University surveyed the mood of villages in Afghanistan in 2011. He found that the villages that were most supportive of the United States were more likely to draw attacks from Taliban forces, but no more likely than other villages to provide information that would help U.S. forces locate improvised explosive devices.

The conclusion that civilian attitudes could have such a strong and direct influence on attacks is a controversial one, but if it holds true upon further investigation, it could give strategists new insight into enemy operations.

Lyall and his team used an algorithm that draws on information from the surveys and derives predictions of attacks. Previous efforts have relied on an area's history of attacks to predict those to come. Lyall's project involved surveys of 2,754 residents in 204 villages.

"Targeted surveys with purpose-built questions are likely to have a higher predictive payoff than large-scale surveys at a fraction of the cost," Lyall says. He is also confident this approach could build prediction models for other types of violence.

The CIA is funding a big-data project with a similar goal, the Political Instability Task Force. That effort uses a large data set, compiled by a program that trawls the internet to gather information. Source