Sign up for our newsletter and get the latest big data news and analysis.

To Crack Down on Crime, Take a Look at the Big Picture

Pete SnellingIn this special guest feature,  Pete Snelling of SAS discusses how law enforcement agencies can yield significant benefit from big data visualization methods. Pete Snelling is a Principal Technical Consultant within the SAS Global Public Security team. In his role, he has also briefed U.K. Members of Parliament on the role of business analytics in policing and worked with domain experts and clients around the globe.

It’s a standard visual in television shows and movies about solving crimes – the big map in the investigation room with push pins showing where each crime occurred and bits of string linking potential suspects with different locations.

That’s very 20th century.

Today, law enforcement agencies can use data visualization technologies that layer crime, traffic, geospatial, weather and other data to see patterns that no amount of push pins and string could ever uncover.

The key to discovering criminal patterns that can break cases wide open is likely buried in the mounds of structured and unstructured data law enforcement agencies gather each day. Being able to surface the data in a meaningful way is critical.

Visualizing the way forward

With data visualization as the result of analysis, tactical and strategic reports can be created rapidly for executive and operational briefings using data from many sources and delivered in multiple formats including web and mobile-based platforms.  This frees up senior officers to spend more time on strategic issues rather than data collection.

Data visualization makes it easier to use traditional (but time-consuming) methods such as the link diagrams which show the relationship between entities; people, objects and locations; and events and each other.  And then it takes it one step further – helping your agency understand correlations, along with possible spatial and temporal relationships between crime, offender and victim.  Is a new way of deploying police having the intended impact? Are there similarities in break-ins that would never have been discovered because what connects them is so seemingly random? Is there a pattern to victims that suggests a preventive intervention? Is there a sudden uptick in individuals with passports from a specific country using one obscure entry port?

Visual data analysis can also be deployed to help law enforcement achieve wider benefits for society. For example, motor vehicle crashes cost society an estimated $300 billion each year and in human terms, the cost is much greater. Law enforcement can use visual analytics to uncover crash causes or identify accident black spots.

Critically too, data visualization drives self-service. IT departments may be feeling the squeeze, but visual analytics is intuitive and supports a self-service approach, enabling law enforcement to deploy it and use it to create their own visual reports without the help of their IT department.

There is of course a catch to avoiding the IT department. The tools used must support the types of data law enforcement agencies collect.   Public security data is often unstructured (witness statements or intelligence reports for instance) and traditional reporting tools struggle with this format.  Your visualization solution should ideally include text analysis tools and the ability to merge text (unstructured) and traditional (structured) data.

It’s also important that the data be managed in an automated manner (captured, transformed and delivered) that doesn’t require manual intervention. Duplicative data must also be handled differently than a mail order catalog company might. As criminals don’t particularly care about offering “one version of the truth” your databases need to capture all the variations of names and addresses that might be out there. Retaining the error, after all, could be important evidence.

Here are some examples of visual analytics in action:

  • One European city was able to demonstrate the reduction in traffic accidents after employing analytics and data visualization to locate accident hot spots and add traffic calming measures. The visual analytic component helped sell the unpopular traffic calming measures to an initially reluctant public.
  • Burglars frequently like to work in one area or against one type of business or residence. Data visualization helps police departments make sense of these connections. Is there an uptick in stolen copper wiring within a six-block radius of a recently paroled burglar? Do apparently random garage break-ins actually share a previously hidden modus operandi?
  • A local U.S. community can easily share data on perpetrators with property managers, allowing them to make decisions about re-signing tenants. In addition, a quick search of any address in the community provides critical information on building structure and access points important in the event of criminal event or fire.
  • By merging data from multiple databases and allowing law enforcement to access info in a mobile format one U.S. state is less likely to let parolees and wanted criminals who encounter law enforcement slip through the cracks. Automated watch lists let law enforcement officials monitor changes in offender status.

Data visualization technology can bring benefits to almost everyone within law enforcement – and the benefits extend into the wider community. The more police are able to see the big picture, the higher the likelihood of crime getting solved quickly. In an era when retailers can anticipate what a shopper wants, the public expects law enforcement to be able to predict and stop crime. And the shift in focus within agencies from solving crime to preventing it is well underway.

The key is using information in a way that isn’t complicated or walled off from other resources. Integration with other agencies, and other types of community records is critical. Black box solutions that work on one piece of the puzzle ramp up the cost without providing the necessary return on investment.

 

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: