Charleston Police to use data mining to fight crime

June 12, 2012

It’s not quite Minority Report but the Charleston Police Department in South Carolina aims to predict and prevent crimes from happening with the help of IBM’s data analytics tools.

To start the CPD wants to focus on reducing the number of burglaries in Charleston which it says often happen at similar times of day and in similar locations.

“The individuals committing these crimes tend to have predictable patterns, and incidents usually take place near their homes or familiar locations. In addition, property crimes are not displaceable crimes, which means the criminals won’t simply move two miles to another location'” the CDP said.

IBM’s predictive analytivcs software aims to uncover hidden patterns and insights from structured and unstructured data to pinpint where and when a crime would be committed. It also promises that its models will help anticipate threats, identify suspicious actors and effectivley allocate resources. Eventually the Charleston police wants to use IBM’s software to localize what it calls criminal hotspots and deploy police officers accordingly to prevent crimes befor they occur.

Charleston has more than 400 police officers who work on evaluating and forecasting crime patterns. It joins cities such as New York, Rochester, Las Vegas, Memphis, Los Angeles, Vancouver or Richmonf that use IBM software to fight crime.

IBM can point to some successes: Richmond, Virginia, which in 2005 was ranked the fifth most dangerous city in the United States dropped to the 99th spot after it began using IBM’s data mining technology to reduce crime.

“Historically, police agencies focused on protecting the community by solving crimes quickly to serve as a deterrent to would-be criminals,” Mark Cleverley, IBM global director of Public Safety solutions said.

No comments so far

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see