Social media can be a real mine of knowledge on a variety of topics. Perhaps soon will also be used by law enforcement agencies to prevent crime. This is at least the intention of a new project implemented by Cardiff University researchers.
They work on a special tool that will allow the police to scan social media, helping identify the focal areas that can lead to hate crimes. The Justice Department-funded project will be based on a special algorithm that scans Twitter by searching for hate speech in specific geographic areas of the United States. The purpose of the algorithm will be to find patterns between the hate speech present in the network and the brutal crime committed in the streets.
If the system is successful, the police will be able to anticipate the outbreaks of violent crimes and hate crimes that could be caused by triggers, such as the death of Keith Lamont Scott, shot dead several days ago by the police.
The learning machine learning system has less offensive words related to hate speech, which are not always directly linked to racism. This will identify cyberbullying in any form.
Los Angeles will experiment with the new system and its task is to allow the police to intervene before the situation gets hot. Of course, before the algorithm becomes fully useful, much work is still left to its creators. You have to learn to identify the phrase with regard to the context, and create a map of the area of the city in which the most common crimes are committed.
Work on the system is planned for at least three years.