Twitter can predict crime before the police do

Twitter has repeatedly proven its usefulness as a platform for the exchange of information in times of social crisis.

Now researchers from Cardiff University have demonstrated the ability of the service to predict a serious accident much faster than the police. In some cases the difference is one hour.

The researchers analyzed data on the riots in London in 2011 and showed that the computer can automatically scan Twitter to detect potentially dangerous events. The system, for example, would be able to tell us about broken Windows and torched cars before this became known to the police of London.

The computer can analyze the rumors in Twitter and determine where the probability of occurrence of the riots above all. The system is also able in real time to report large clusters of people.

The researchers used a number of machine learning algorithms to analyze 1.6 million tweets. The system took into account the time and place of publication of the messages and their contents. The algorithms found the scene faster than the police in all but two cases.

One of the study’s authors, Dr. Pete Burnap (Pete Burnap), believes that technology monitoring Twitter can help you deal with future riots. “Earlier we used to process Twitter data using machine learning and natural language systems for a better understanding of online disabilities such as the dissemination of hate speech and cyber hate”, he said.

“In this study, we show that social networks become a starting point for communication about everyday events, including social unrest and terrestrial criminal activities — continued Burnap. We could never replace traditional police ground resources, but we have demonstrated that this research can improve existing data collection systems and allow to create new technologies to support the established methods of policing”.

© 2017, paradox. All rights reserved.