U.S. Government to Use Social Media to Predict 'Societal Events'

The billions of words and millions of photos people voluntarily post to social media websites like Twitter and personal blogs may soon help the government predict events like protests, riots and even terrorist attacks. The Intelligence Advanced Research Project, a U.S. government agency that invests in high-risk/high-payoff research programs, announced in August that they are searching for ways to mine data to predict events that include political and humanitarian crisis, mass violence, mass migrations, disease outbreaks, economic instability, resource shortages and responses to natural disasters, IARPA wrote in a statement.

The data mining initiative from the U.S. intelligence agency is called the Open Source Indicators Program. The program reflects the growing trend of decision making based on advances in computational intelligence. IB Times reported earlier in November that public health experts are using open-source data in a similar fashion, monitoring so-called trending topics for signs of disease outbreaks.

National security and public health are just a couple of the fields for which computational intelligence can be applied, Nature News reported in October. Corporations can use open source data to predict which products will sell and which movies will tank.

Peter Gloom, a computer scientist at the Center for Collective Intelligence at the Massachusetts Institute of Technology, told Nature News, We're up to 90 percent accuracy for predicting box-office returns.