Inside Report: BRAVE crime fight focuses on algorithm, mapping

In a traditional sense, brokers manage money.

But to Baton Rouge Violence Area Elimination project researchers, “brokers” are people with unique connections to multiple gangs across East Baton Rouge Parish.

“These are important people because they carry a lot of information,” said Tracey Rizzuto, a BRAVE researcher and LSU associate professor who analyzes crime data.

The fun fact was one of many divulged by researchers during a recent presentation of BRAVE findings and methodology at LSU.

During the presentation, Rizzuto pointed to a large, illuminated screen peppered with dots of various colors. Some of the symbols represented gangs, while various dots symbolized people who have connections among multiple gangs and therefore might be valuable to law enforcement agencies, Rizzuto said.

“We’re looking to get at the root of the causes (of violence) within these communities,” she said, echoing officials who have explained BRAVE as a community-policing method aimed at reducing violent crime, particularly among young men in Baton Rouge’s crime hotspots.

Rizzuto, who studies social networks and group dynamics, said computing tools help researchers figure out which gangs are growing, which gangs are shrinking, and which people associate with different gangs.

Researchers then share that information with law enforcement agencies that can sometimes use the data for real-time assistance.

For example, researchers have compiled a fluid list known as the “hot list,” which identifies about 200 of the most violence-prone youths in the parish.

Baton Rouge Police Chief Carl Dabadie said patrol officers have access to that list from the computers in their squad cars.

So if an officer comes across a person on that list, the officer knows whom he or she is dealing with, Dabadie said.

Anthony Reed, an LSU graduate student who works alongside criminologist Ed Shihadeh, said that researchers developed an algorithm that quantifies a person’s propensity to get involved with criminal behavior. The algorithm, based partly upon past criminal histories and known gang associations, spits out a composite score, allowing researchers to rank the most dangerous people.

The formula also focuses on people who weigh most heavily on limited police resources, such as those who commit crimes late at night or in densely populated areas, Reed said.

Shaun Williams, an LSU geography and anthropology graduate student who works on the BRAVE team, has helped develop maps that show where crime hotspots have shifted over the past year. The maps focus on a three-layer umbrella of corridors, neighborhoods and people, Williams said.

In one map, Williams showed how data collected connects some East Baton Rouge Parish gangs to gangs across the river or as far away as New Orleans.

Cecile Guin, BRAVE’s lead researcher and principal grant writer, said she hopes to soon submit a grant application that would expand BRAVE’s services from the 70805 and 70802 ZIP codes to the 70810 ZIP code covering the Gardere area. While researchers found violent crime dropped in many pockets of Gardere from 2012 to 2013, the neighborhood remains one of the most violent areas in the parish.

As long as the grant money keeps flowing, researchers plan to continue sharpening law enforcement’s data-driven toolset.

Ben Wallace covers law enforcement for The Advocate. He can be reached at bwallace@theadvocate.com or on Twitter at @BenWallace.