Crime predicting software helps police

There is a new king of software that has been recently developed, a crime prediction software, which is supposed to reduce not only the murder rates, but the rates of many other crimes as well.

The software has been developed by Richard Berk, a professor at the University of Pennsylvania, and it is already used in Baltimore and Philadelphia to help police figure out which ex cons on probation or parole are most likely to murder and to be murdered. The latest version, which has been implemented in D.C., has options permitting to identify the individuals which are most likely to commit other crimes than murder. If after a period of time Berk’s software proves to be successful, it may influence sentencing recommendations and bail amounts. A person that goes on probation or parole is supervised by an officer, who uses the person’s criminal record, and their good judgment, to determine the level of supervision which is needed in every particular case, with every ex convict. This software is supposed to replace the calculations and observations of the police officer and determine the level of supervision which is indicated in every case.

Murders are the most rare crimes: there is one murder for every 100,000 people, rising to a rate of one in 100 among high risk groups. Trying to predict such a rare event is difficult and many researchers thought it was impossible. But the new advances in computer technology are able to do what everyone thought to be impossible. Several years ago, the researchers started creating a database with more than 60,000 various crimes, including murders and using an algorithm they can now predict which people are much more likely to commit homicide when paroled or probated. Instead of finding one murderer in 100, the UPenn researchers are able to identify eight future murderers out of 100. Berk’s software takes into consideration two dozen variables, from criminal record to geographic location and is able to predict even the type of crime and the age at which that crime will be committed. Berk says that people generally presume that if someone murdered then they will murder again in the future, but there are particular cases: if the individual committed armed robbery at age 14 that’s a good prediction, but if they committed the same crime at age 30 that does mot mean very much.

Baltimore and Philadelphia are already using Berk’s software to determine how much supervision parolees they should have and Washington D.C. is planning on using the algorithm to determine smaller crimes as well. If the tests go well, Berk says that the program could help set bail amounts and suggest sentencing recommendations. Scientifically, Berk’s results are impressive, but predicting rare events like murder is extremely difficult. By considering a group of people as high risk, and watching them with increased vigilance, there should be fewer murders and other crimes. Inmate rights advocates argue that this is a form of harassment, because no one can be sure that they will commit a crime again in the future.