Major advancements in machine learning and artificial intelligence have pervaded almost every facet of daily life, and crime is no exception. People and governments have been obsessed with how AI can be used to deter crime for decades—as early as 1956, sci-fi author Philip Dick wrote “The Minority Report” among his personal Cold War fears.
Some models have found success, like ShotSpotter, but certainly nothing like what was described in Dick’s book or the Steven Spielberg film adaptation. Still, there’s something to be said for utilizing advanced technology when possible.
Scientists from the University of Chicago have done just that with the development of a new algorithm that predicts crime a week in advance with up to 90% accuracy.
“We created a digital twin of urban environments. If you feed it data from happened in the past, it will tell you what's going to happen in future. It's not magical, there are limitations, but we validated it and it works really well,” said Ishanu Chattopadhyay, professor of medicine at UChicago. “Now you can use this as a simulation tool to see what happens if crime goes up in one area of the city, or there is increased enforcement in another area. If you apply all these different variables, you can see how the systems evolves in response.”
How it works
The novelty of the new model is that it forecasts crime by learning patterns in time and geography. In contrast, most other previous efforts relied on an epidemic or seismic approach, where crime was depicted as emerging in hotspots that spread to surrounding areas. The UChicago scientists say the epidemic approach misses out on the complex social environment of cities, and doesn’t consider the relationship between crime and the effects of police enforcement.
“Spatial models ignore the natural topology of the city,” said sociologist and co-author James Evans, professor at UChicago and the Santa Fe Institute. “Transportation networks respect streets, walkways, train and bus lines. Communication networks respect areas of similar socio-economic background. Our model enables discovery of these connections.”
The UChicago model isolates crime by looking at the time and spatial coordinates of discrete events and detecting patterns to predict future events. It divides the city into spatial tiles roughly 1,000 feet across and predicts crime within these areas instead of relying on traditional neighborhood or political boundaries.
The tool was tested and validated using historical data from the City of Chicago around two categories of reported events: violent crimes (homicides, assaults and batteries) and property crimes (burglaries, thefts, and motor vehicle thefts). The team chose to incorporate that specific data for two reasons: 1) these crimes are most likely to be reported to police in urban areas with historical distrust and cooperation with law enforcement, and 2) these crimes are less prone to enforcement bias.
The study results, published in Nature Human Behavior, showed the model can predict future crimes in Chicago one week in advance with about 90% accuracy. Importantly, the model performed just as well when the researchers inputted data from seven other U.S. cities—Atlanta, Austin, Detroit, Los Angeles, Philadelphia, Portland and San Francisco.
In a separate model, the team also studied police response to crime by analyzing the number of arrests following incidents and comparing those rates among neighborhoods with different socioeconomic status. They report that crime in wealthier areas resulted in more arrests, while arrests in disadvantaged neighborhoods dropped. Crime in poor neighborhoods didn’t lead to more arrests, however, suggesting bias.
“What we’re seeing is that when you stress the system, it requires more resources to arrest more people in response to crime in a wealthy area and draws police resources away from lower socioeconomic status areas,” explained Chattopadhyay.
While the model proved to be powerful and accurate, Chattopadhyay cautioned that that doesn’t mean it should be used to direct law enforcement to specific areas to prevent crime. Rather, he says, it should be just as another “tool” to help urban police address crime.