Big data takes a shot at crime in Saskatchewan
But now, a partnership between a team of researchers from the University of Regina’s Department of Computer Science and ISM Canada is creating new tools using “big data” that can help to tackle crime on the streets using information from the virtual world.
Seeking a big data solution
As a specialist in information technology storage and analytics, ISM Canada provides important support services to many large corporations and governments, including the Saskatchewan Department of Justice. ISM was wondering if connections could be made using millions of pieces of data in ways that might help them to create new crime reduction initiatives in the province.
Recognizing the challenging volume of data involved, ISM partnered with Professor Howard Hamilton at the University of Regina and a team of six researchers for their expertise in the latest data science techniques.
Mehdi Sadeqi, a postdoctoral fellow and the team leader, explained their approach: “One of the really interesting things about crime information is that we can use it to make connections, and then build predictions. Take graffiti for example. We could map out all the areas that have had increased incidents of graffiti and overlay it with information about other crimes in that location, like break-and-enters. If we notice a connection between the two, then we can use that knowledge help the graffiti taggers avoid criminal lifestyles that might lead them to committing break-and-enters. In this way, we’re using big data to prevent criminals from re-offending, and hopefully taking them out of the cycle of crime altogether.”
The team of researchers draws from a diversity of expertise including data mining and visualization, privacy protection, and social network analysis. Masters of Computer Science student Credell Simeon details her role: “For my part of the project, I added in so-called ‘sentiment analysis’ from social networks. We can tell if a person expresses intent to commit crimes using publically available data like open tweets or posts to different social media sites. When we combine this with the other information available, we can get a much richer picture of how and where crimes might take place.”
Once complete, the data mining and visualization system will be an important tool to support crime-reduction programs. The knowledge gleaned could help support offenders from becoming involved in the cycle of crime, potentially leading to a reduction in the province’s crime rate in the long term.
A chance to sharpen their skills
Both Mehdi and Credell expressed pride for participating in such an important and large-scale initiative.
“This project has really helped me to expand my knowledge of both the theoretical and practical aspects of my field,” says Mehdi. “Working with the team at ISM has helped me to understand what it’s like to work in industry and to meet business demands.
“And as the project leader, I am learning how to motivate my team and to work to everyone’s strengths so that we can present the best product to our client. In the end, this project is giving me the freedom to choose between academia and industry because I have now seen both sides first-hand.”
For Credell, the project has given her insight into her possible future. “As an international student, Mitacs Accelerate gave me valuable experience that will help me as I begin my job hunt in Canada. I have a better sense of what companies want and how I can apply my skills to help them. I am very grateful to have had this experience.”
Photo at right: The Mitacs Accelerate project team, supervised by Professor Howard Hamilton (at centre).
Mitacs would like to thank the Government of Canada along with Alberta Innovates, the Government of British Columbia, Research Manitoba, the Government of New Brunswick, the Research & Development Corporation of Newfoundland and Labrador, the Government of Nova Scotia, Innovation PEI, the Government of Quebec, and the Government of Saskatchewan for their support of Mitacs Accelerate.
Photo at top courtesy Flickr user John Backstrand. Used under Creative Commons Attribution 2.0 licence.