An eight-year-old boy goes missing in Saskatoon, Sask. Investigators scramble for leads — the child could be anywhere. Best case scenario: he's run away and will come home soon. Worst case scenario: he's been abducted.
So how do police know where to look? With the help of a new computer lab, the Saskatoon Police Service (SPS) is hoping to predict, prevent and find missing children — using numbers.
"I'm not a mathematician, I don't understand algorithms," says SPS Deputy Chief Jeff Bent, who has been a police officer for more than 30 years. "But we have to start recognizing the potential of predictive analytics for solving police problems and crime."
That's why the SPS, along with the provincial Ministry of Justice, employed the help of computer scientists and mathematicians from the University of Saskatchewan. The team of researchers will tackle police problems that are rooted in data, to try to prevent crimes before they happen. Their first project will be finding missing children, as part of the Missing Persons Project.
"Some of these kids are going to end up in the sex trade, and that goes across borders and municipalities," says Brian Rector, executive director, Ministry of Justice. "To be able to pull all that data together — it could save lives."
Starting this year, the Saskatoon Police Predictive Analytics Lab (SPPAL) will take data provided by the RCMP, municipal police services in Saskatchewan and child services to map out the risk factors, trends and behaviour patterns of the province's missing youth. The lab will assess each child's risk of going missing, and use that data to help find them if they disappear.
"We're doing more than law enforcement, we're going out there and trying to stop the crimes from happening to begin with," says Bent. "We hope this lab can help us intervene earlier in the lives of these vulnerable people who may otherwise become victims or active criminals themselves."
By the numbers
Housed at Saskatoon police headquarters, the new lab looks small and unassuming. But the nondescript room is filled with humming desktops and tapping computer scientists who will sift through thousands of data files to find trends and patterns among victims and criminals.
Most police agencies already use data and statistics to map out crimes, using a tactic called hot-spotting to find the most crime-ridden areas.
But now, many police organizations, including the SPS, are stepping up their number-crunching game. They're moving away from the traditional statistical methods, focusing instead on probabilities and large-scale patterns solved with algorithms.
"This type of data analysis has been applied to many different organizations, so why not police?" says Daniel Anvari, a researcher from the University of Saskatchewan.
This form of data mining, called predictive analytics, is used for everything from calculating credit scores, to generating tailored deals for loyal customers. When used in policing, predictive analytics can help track individuals and incidents to prevent and solve crimes.
"This will help police make better decisions," says Anvari. "The algorithm isn't going to make any decisions itself, but it may highlight something that could be missed from just looking at the data with our eyes. There might be hidden patterns or correlations that we aren't aware of."
New and old
Police have always used the latest technology and hardware to improve their front-line operations, from radar guns to robots. Predictive software is just one more tool to boost police efficiency in an increasingly digital world.
The SPS plans to combine this new software technique with old practices: preventative policing. Police officers will use the data to pinpoint the highest risk individuals, offering them social services and healthcare before their situation escalates.
"There are many police related issues that aren't just 'catch the bad guy'. Many of them are social, not just criminal, and can be prevented beforehand," says Stephen Wormith, a professor at the University of Saskatchewan who is leading the lab. Wormith is also the Director of the Centre of Forensic Behavioural Science and Justice Studies, which is participating in the SPPAL's Missing Persons Project.
Looking forward, Wormith says the team is "only limited by our imagination and our creativity" for the types of projects they can tackle — a prospect that excites Bent.
"We don't know exactly what we're going to find, maybe it's just going to confirm things we know from experience and police intuition," says Bent. "But we all believe there is great potential. Maybe there will be something that's revolutionary that no one has discovered before."