Artificial system

Researchers develop an algorithm to predict crime a week in advance

Social scientists from the University of Chicago have developed an algorithm that can predict crime in urban areas up to a week in advance, Bloomberg reported Thursday.

Over the past few years, there has been a huge increase in the use of algorithms around us. Whether it’s predicting the weather, driving cars, making shopping recommendations or finding cures for diseases, algorithms are at work everywhere. It would hardly be a surprise if they weren’t used to fighting crime.

Prior to the Olympics, Tokyo police were looking to implement artificial intelligence (AI)-based technology to predict crimes before they could happen. If it seems that we live in a Minority report the future already, the fact is that we already have been for almost a decade now.

Chicago Crime and Victimization Risk Model

According to the Bloomberg report, the Chicago Police Department implemented the Crime Risk and Victimization Model in 2012 with the help of academic researchers. The model used factors such as age and arrest history to prepare a list of potential abusers and their victims and even assigned a score to those listed to help law enforcement confer the urgency to follow the alleged perpetrator and his victim.

The concept may sound interesting, but the actual application was questionable. As investigations later showed, nearly half of the alleged perpetrators on the list had never been charged with unlawful possession of weapons, while others had never been charged with serious offenses. previously. A technology review report in 2019 detailed how the risk assessment algorithms that determined whether or not an individual should be sent to prison were trained on historically biased data.

So when researchers at the University of Chicago, led by Assistant Professor Ishanu Chattopadhyay, tried to build their algorithm, they wanted to avoid the mistakes of the past.

How does the new algorithm work?

The algorithm divides a city into 1,000 square foot tiles and uses historical violent and property crime data to predict future events. The researchers told Bloomberg that their model is different from other such algorithmic predictions because others view crime as emerging from hotspots and spreading to other areas.

However, such approaches, according to the researchers, do not take into account the complex social environment of cities and are also biased by the surveillance used by the state for law enforcement. Instead, the algorithm used analyzes previous crime reports taking into account many other factors and then predicted the likelihood of crime in Chicago with 90% accuracy. The model was also used to predict crime in eight different cities in the United States, which included big names like Los Angeles, Atlanta and Philadelphia, and performed well in those scenarios as well, Bloomberg said in its report.

The algorithm and scientific details were published in the journal Natural human behavior.


Police efforts to thwart crime are generally based on reports of criminal offences, which implicitly manifest a complex relationship between crime, police services and society. As a result, crime prediction and predictive policing have generated controversy, with the latest artificial intelligence-based algorithms producing limited insight into the social system of crime. Here, we show that while predictive models can enhance state power through criminal surveillance, they also enable state surveillance by tracing systemic biases in crime suppression. We introduce a stochastic inference algorithm that predicts crime by learning spatiotemporal dependencies from event reports, with a mean area under the receiver operating characteristic curve of approximately 90% in Chicago for predicted crimes. per week at approximately 1,000 feet. Such predictions allow us to study disruptions in crime patterns that suggest that the response to increased crime is biased by neighborhood socioeconomic status, draining political resources from socioeconomically disadvantaged areas, as demonstrated in eight major US cities.