Purdue University students have developed a software program that uses Big Data to help police target where crime could happen.


Purdue’s software uses algorithms to predict criminal activity.

The program, called “Valet,” took more than four years to develop. It’s already being deployed to police agencies in several states.

Lafayette Police Chief Patrick Flannelly says the program gives police real-time data to work with and can help shrinking departments do more with less.

The program places red spots on a map to indicate where crimes have been committed and the nature of those crimes. It uses police data, court records and social media to track patterns that police can use to beef up patrols at key times.

It also uses algorithms to forecast future criminal activity, with 95 percent certainty.

“It’s basically a way for them to get a better understanding of a situation they’re about to find when they knock on that door,” said Purdue Professor David Ebert, who led the team of researchers at Purdue’s Homeland Security Center of Excellence.

See Page 2 for more on what police learn from Big Data used to fight crime.