Developers: | University of Chicago |
Date of the premiere of the system: | August 2022 |
Branches: | Law |
2022: Announcement of an AI system predicting crimes with 90% accuracy
On August 19, 2022, it became known that specialists from the University of Chicago created an algorithm for predicting crime by analyzing natural phenomena and predicting possible future crimes in one week with an accuracy of more than 90%.
The algorithm was successfully tested and validated using historical data for the City of Chicago for two broad categories of recorded events. These are crimes of a violent and property nature. Violent crimes include murder, assault and battery. Property crimes include burglary and theft of vehicles.
The use of this data is due to the fact that they are most likely to be reported to the police in urban areas, where there has historically been distrust and lack of cooperation with law enforcement agencies.
The algorithm analyzes the pattern of events in time and geography based on open data on violent and property crimes. He divides the city into squares of about 300 meters each and calculates crime in these areas.
The model demonstrated a similar result on data from seven other cities: USA Atlanta, Austin, Detroit, Los Angeles, Philadelphia, Portland and San Francisco.
One of the lead authors of the study, Ishanu Chattopadhyay, emphasizes that the high efficiency of the algorithm does not mean that it can be used to guide law enforcement policies. According to Chattopadhyay, police departments, for example, should not use it to pre-emptively survey areas to prevent crime.
Instead, it needs to be included in the city's policy toolkit and crime policing strategies.
We have developed a digital dual model of the urban environment. If you give the algorithm information about what happened in the past, it will show you what will happen in the future, "says Chattopadhyay. This is not a magical invention, there are limitations, but we have experienced the operation of the algorithm in practice, and it does an excellent job. |
This algorithm should be applied as a modeling tool to see what happens if the crime rate rises in one area of the city, or if law enforcement is strengthened in another area. If you apply certain solutions, it will be possible to observe how the systems will develop in response, "he added.[1] |