Memphis cops use predictive analytics to fight crime
Predictive models help them pinpoint crime hot-spots and dispatch more officers
It can sometimes be hard to pinpoint the return on investment in a particular technology, but police officials in Memphis, Tennessee, say they are getting major results from using predictive analytics technology to fight crime.
Several years ago, police there began developing a program called Blue CRUSH (Criminal Reduction Utilizing Statistical History) along with the University of Memphis' Department of Criminology and Criminal Justice.
It uses predictive analytics software from IBM's SPSS division to crunch information showing what types of crimes are occurring, where and when. Police officials use the resulting models to dispatch more resources to the neighborhoods that appear to need the most help. Blue CRUSH also uses a mapping program to provide visual representations of crime patterns.
The software models have changed the game for policing in Memphis, said John F. Williams, crime analyst unit manager. "This is more of a proactive tool than reacting after crimes have occurred. This pretty much puts [officers] in the area at the time they're being committed."
One operation in January focused on the city's Hollywood-Springdale area. It netted more than 50 arrests and crime there has fallen 36.8%, according to a statement.
Overall, police officials have seen a return on investment of 863%, with an average annual benefit of more than $7 million, according to a case study done by Nucleus Research.
"MPD has seen a reduction in violent crime, property crime, and part 1 crimes (which include violent and property crimes) by an average of 15.8% without a corresponding increase in officers -- while expanding its overall geographical coverage," Nucleus' report states.
Blue CRUSH is also boosting police officers' morale, as they are now making more "quality arrests" of people who are repeat offenders, Williams said.
Meanwhile, IBM is betting big on its overall BI (business intelligence) strategy, forming a consulting unit with 5,000 employees and making nearly 20 acquisitions in support of it.
As a subset of the practice, predictive analytics is making significant progress, according to Big Blue. It is currently working with a quarter-million clients on such projects, IBM said.
Predictive analytics are "coming into the core" of companies' BI strategies, according to Forrester Research analyst James Kobielus.
"It's less and less regarded as a silo for the PhDs," he said. "It's being regarded strategically as a decision support tool that needs to be leveraged across a broader range of products than ever before. [Companies are] coming to the likes of an IBM ... saying we have the need for your expertise now."
Expect to see predictive analytics dovetail more closely with CRM (customer relationship management) strategies, particularly in analyzing social networks and social media for insights into consumer activity, Kobielus added.
"A lot of this is converging with CRM. You can expect to see CRM vendors embed predictive models into their overall product portfolio," he said.
It's also likely that CRM vendors will acquire predictive analytics vendors as well as partner with companies like SAS Institute and IBM, according to Kobielus. "All of those guys, big and small, they're going to need to add this to their applications to stay in the running in a world where more interactions with customers are in social media."
Chris Kanaracus covers enterprise software and general technology breaking news for The IDG News Service. Chris's e-mail address is Chris_Kanaracus@idg.com