1 Comment

Summary:

New research from the University of Michigan highlights the potential for tackling tough problems like crime by getting creative with data, so we can fight the disease instead of the symptoms.

I’m taking a break from writing about the nitty gritty of Hadoop to highlight yet another promising use case for big data — fighting crime. A new study by University of Michigan researchers details a method for using — according to a university press release — “high-powered computers and loads of data” to help police target neighborhoods most susceptible to high crime rates.

However, it’s not the idea of using data to predict or even solve crime that excites me. We’ve seen those ideas floated before by IBM and even other researchers, and some police departments are already using them. Rather, I’m intrigued by the possibility that this new research could actually help prevent crime by uncovering causes of crime that might have been ignored previously or were just too latent to actually take into account.

Source: University of Michigan

Source: University of Michigan

The researchers used myriad data sources — everything from demographic data to drug offenses to the types of alcohol served nearby — in order to create a crime heatmap of Boston. They even considered how the attributes of adjacent neighborhoods affect can affect crime rates in their neighbors. As they add in more data, the researchers think they’ll be able to get a better idea of how those additional variables affect crime rates.

The results, at least as described in the press release (the full paper is available for purchase), are somewhat revelatory, but you can read them yourself.

The bigger picture is one I outlined recently while discussing the promise of big data with regard to gun violence. When we get creative and use the countless data sources now available to us, combined with unprecedented computing capabilities, we can start to analyze things in new ways and see relationships we might not have seen before. In the case of crime broadly or gun crime specifically, there might be geographic, socio-economic or public policy factors that could help governments fight the disease rather than just the symptoms.

Structure:Data: Put data to work. 60+ big data experts speaking. March 20-21, 2013, New York City. Register now.Of course, the areas that stand to benefit from big data techniques go far beyond crime. Among the ones we’ll cover at Structure: Data (March 20-21 in New York) alone are health care, medical research, personal finance, national security and commerce. The way I see it, the technology keeps improving and the data keeps proliferating, so it’s up to us to figure out how to use them to solve some of our thorniest business and social problems.

Feature image courtesy of Shutterstock user J.D.S.

  1. So could this data be interpolated to retroactively predict some of the recent horrific gun violence in the society that had taken place? Prove it ! Predict a crime that will happen in the next few months for example …othwise this is all hot air ..

    Share

Comments have been disabled for this post