Cities Turn to Analytics for Targeting Habitual Criminals

In the coming months, the Rochester, Minn., Police Department will implement IBM identity analytics software to assist with “intelligence-led” policing methods, the company announced Wednesday, Jan. 25.

The data intelligence package will focus on identifying repeat offenders rather than predicting crime in the classic sense. In other words, the software is identity analytics, not predictive analytics.

Police Department Lt. Tim Heroff said a large percentage of crime is committed by a relatively small number of individuals. If the most prolific offenders are identified, the thinking is that the police department will be able to more efficiently deploy its manpower and resources.

“In other words, crime analysis has a certain temporal and spatial capacity that’s important in identifying hot spots,” Heroff said. “With intelligence-led policing, we are interested in prioritizing our resources so that we’re looking at the more serious and prolific offenders.”

Rochester, a small town located about an hour and a half south of Minneapolis, isn’t a high crime area, but Heroff said the department is committed to disrupting the criminal activity that does take place. Heroff said implementing the identity analytics software will represent a major paradigm shift for the department’s policing methods. A year and a half ago, the Rochester Police Department adopted the “intelligence-led” policing policy.

The department is slated to implement IBM’s InfoSphere Identity Insight software, which the city’s law enforcement will use to identify crime patterns. According to the company’s description of the software suite, the product accumulates a self-correcting history of an individual over time and identifies associations and networks of people. Law enforcement agencies across the U.S. already are using similar solutions to combat waste, fraud and abuse.

Total costs involved with the project’s implementation are about $820,000, Heroff said.

Mark Cleverley, IBM’s director of global public safety, said the software will help identify duplications in criminal records, which helps law enforcement target who is committing crimes. New police records can be associated with older records, and the software makes connections that otherwise would not occur.

Cleverley said for law enforcement, identity analytics can be a prerequisite for using other types of analytics to help fight crime.

Rochester, Minn., is following a path blazed more than decade ago by metropolitan areas  such as New York City, where officials installed CompStat, an analytics-based policing method that officials credited with lowering crime rates.

In 2010, the Chicago Police Department implemented predictive analytics software to determine where crime is likely to occur before it happens

Rochester, Minn., is another example of law enforcement’s usage of analytics continues to evolve.