The New York City Police Department is using a new software system called Patternizr. The Washington Post reports that policemen can help search hundreds of thousands of case files.
According to the report, the software was developed in-house and allows analysts to search a wide range of files for patterns or similar crimes. Previously, we had to go through physical files. As an example, a police officer used this system to connect two criminals using a syringe to steal training at two different home depots in New York City. Crime analyst Rebecca Shutt, who solved the case states that the system "complained in other areas I would not have known."
This Minority Report- is not a system for analyzing AI through CCTV images as well as a system for predicting where crime will occur. Rather, it is a system that retrieves patterns from the NYPD's database and allows investigators to search across a much wider range of data during the investigation. This system can bring additional sources of information across the NYPD, making it difficult to find patterns of crime that can happen elsewhere.
NYPD announced that this software was released for the first time in issue INFORMS Journal on Applied Analytics although this department released the software in 2016. According to Evan Levine, assistant professor of data analysis at NYPD and executive director of Alex Chohlas-Wood, the department spent two years in software development and claimed that NYPD was the first to use such a system in the United States.
Chohlas-Wood and Levine used a previously identified pattern of 10 years to train the system in and in the test they correctly reproduced the old crime patterns that correspond to one-third of the time " Status states that as a precaution against racial discrimination, the system does not consider the race of the suspect in the search process.
The results seem to help investigators reduce some of the work required and partially automate the manual processes so far.