Can we predict where mass atrocities and other forms of political violence are likely to occur based on known variables such as infant mortality rates (which happen to be good indicators of state failure) or instances of hate speech on Twitter? We're getting there.
An article in today's New York Times by UN correspondent Somini Sengupta describes a number of ongoing projects that use content analysis from newspaper archives, state-level data (on, for example, defense budgets and infant mortality), social media postings, and more in an effort to predict various forms of violent conflict. Efforts to turn data into foreknowledge--or at least better risk assessment--have been funded by the U.S. intelligence community for years. The Political Instability Task Force, based in the Center for Global Policy at George Mason University, is the center of CIA-funded open-source analysis.
 |
| Choeung Ek, Cambodia |