The Chicago Police Department is developing a predictive analytics system with the help of a local university.
The Chicago Police Department is teaming with a local university to develop a system that predicts where crime will occur.
According to the Chicago Sun-Times, the city's police department, the Illinois Institute of Technology and Rand Corp. are working together on the system, which predicts crime hot spots. With the data, the police can deploy resources more intelligently into the affected communities.
Funded by a $200,000 grant from the National Institute of Justice, the stakeholders are tracking every incident that's linked to a known gang member in order to build the analytics engine. Chicago Police Superintendant Jody Weis revealed the city's new crime-fighting strategy at a press conference Sunday, Aug. 8. The initiative was launched in April, he said. The Chicago Police Department also launched an analysis group for the program.
The policing approach, called predictive analytics, has gained momentum in recent years as law enforcement agencies have recognized that some types of crime follow patterns that can be predicted by software.
In one example, since 2006 the Memphis Police Department (MPD) has used IBM predictive analytics technology, which officials credit with helping reduce crime by 30 percent. To manage and compare crime data from years past, the MPD created Blue CRUSH (Criminal Reduction Utilizing Statistical History), built in partnership with the University of Memphis' Department of Criminology and Criminal Justice. Crime analysts plug the records from Blue CRUSH into the predictive tool.
A few years ago, in another program, 25 law enforcement agencies, including Utah and San Jose, Calif., began testing a crime analytics tool called Command Central from vendor CrimeReports. Researchers at the University of California, Los Angeles developed a math-based simulation model that analyzes how different crime hot spots respond to increased policing. The Los Angeles Police Department is currently using this model.