AI can now tell if you’re lying….well almost. Police stations in Spain are using artificial intelligence to help identify false robbery reports based solely on their text.
Created by computer scientists from Cardiff and Charles III Universities in Madrid, VeriPol was made specifically for work in law enforcement. It combines Natural Language Processing and Machine Learning methods to judge robbery reports probability of being false. Veripol was trained using 1000 robbery reports including those that were confirmed to be false.
The program uses algorithms to identify various characteristics in the written reports including verbs, adjectives, and punctuations. The artificial intelligence then looked for certain patterns to identify the false reports. According to studies by Cardiff University, false police reports are more likely to be shorter, focused on the stolen property rather than the robbery itself, have few details about the attacker or the robbery, and lack witnesses.
Veripol Artificial Intelligence Can Identify False Police Reports
In a 2017 test study conducted in the cities of Murcia and Malaga, the program performed with flying colors. In reports that Veripol marked as having a high probability of being false, 83% of those cases were closed after the claimants were questioned further. The program identified 64 false reports in one week. Using the same set of data, Veripol had an accuracy of more than 91% while it’s human counterparts were only 76% successful in identifying the false reports.
AI Can Help Police Save Time
Scientists believe the tool can act as a partner to help save police time during investigations while also deterring people from filing fake reports.
“Police officers across Spain are now using VeriPol and integrating it into their working practices. Ultimately we hope that by showing that automatic detection is possible it will deter people from lying to the police in the first instance.” – said co-author of the study Dr. Jose Camacho-Collados, from Cardiff University’s School of Computer Science and Informatics.