IE 11 is not supported. For an optimal experience visit our site on another browser.

Gangs' pecking order revealed by Army software

ORCA
ORCA process reducing a group of individuals (A) to its most influential ones, based on several metrics.Damon Paulo / West Point Military Academy

Remember those pegboards covered in photos and string that adorn the gang-busting headquarters of Hollywood (and real) cops? That time-honored tactic getting the automation treatment with a new Army-developed tool that uses arrest information to make a map of criminal organizations.

Facebook, LinkedIn and other sites aren't the only social networks out there from which to gather information. We form connections based on where we live, who we hang out with and what we do. In the case of an ordinary citizen, that might mean your network comprises a neighborhood and a hobby group — but for a gang member, things like drug choice, tattoos, and criminal history come into play.

Such connections are invaluable for law enforcement if they are to understand the organization of gangs in addition to using dangerous methods like direct surveillance or going undercover.

Enter the Army. It created a tool for use in Afghanistan and other war-torn areas that collates information about people known to be related in some way to enemy forces, spitting out networks based on statistical connections between crimes, locales, relationships to central figures, and so on. Analysts at West Point Military Academy thought it might be useful for police trying to suss out connections in gangs and organized crime, so a military team started adapting it.

Their results are summarized in a new paper, "Social Network Intelligence Analysis to Combat Street Gang Violence," which you can read here on Arxiv.

The tool, called ORCA ("Organizational, Relationship, and Contact Analyzer"), took in 5,418 arrests, comprising 1,468 individuals belonging to 18 gangs. It created a database of connections in seconds, identifying major players, people who live at the borders between gangs, and people who may not say they're in a gang but very likely are.

Some of it might be new information, for example, likely gang members. And some might back up what police on the street already know, such as "corner crews" that deal drugs regularly in one place. But apparently it's all useful: the creators write in the paper that they're working with a major metropolitan police force to deploy the system for real, rather than experimental, use.

Devin Coldewey is a contributing writer for NBC News Digital. His personal website is coldewey.cc.