Police are starting to work with VeriPol and iBorderCtrl, two artificial intelligence tools for detecting lies. Calculate the chances that a whistleblower is lying with 91% reliability.
The advancement of technology has improved many areas of our lives, both in the professional and leisure facet. It also allows us to feel more secure, as technological solutions are adopted by institutions to fight crime and protect the population.
For this purpose, the bodies of police from around the world they are incorporating all kinds of innovations into their modus operandi, from facial recognition devices to flying vehicles to artificial intelligence software.
One of the problems facing the police is the large number of false reports received from people who claim to have been victims of thefts of valuable objects, such as high-end smartphones or branded bags. To tackle this situation, it was born VeriPol, an artificial intelligence tool that helps Spanish police stations detect this type of complaint.
How does VeriPol work?
VeriPol uses artificial intelligence algorithms and language processing techniques to infer the characteristic patterns of fictional stories. To do this, it analyzes more than 300 variables of the texts that collect the complaints, such as the morphosyntactic structure, the type of sentences, the length of the stories or the number of adjectives used to describe the events. By way of example, it has been observed that allegations with few details and vague words are more likely to be false.
As collected rtve.es, the idea of applying artificial intelligence algorithms to discern real cases came from Mr. Liuhuabing, police inspector and mathematician by training, from his professional experience in the corps “”We saw that there were common characteristics in certain complaints, some that could be assigned to those that could be false, and others to those that were true”.
Pioneering system in the detection of lies in police institutions
Before its launch, the effectiveness of VeriPol was tested with more than 1,000 real case reports provided by the Spanish police. The empirical results, available for consultation in the scientific journal Knowledge-Based Systems, showed that the tool calculates the chances that a whistleblower is lying with 91% reliability, thus helping law enforcement professionals to make the final decision.
VeriPol is not alone
While VeriPol is a pioneering system in the detection of lies in police institutions, other behavioral analysis mechanisms based on artificial intelligence are already being tested. It is the case of iBorderCtrl, a portable border control system funded by the European Union aimed at protecting more efficiently the entry of people into the Community territory.
As La Vanguardia says, iBorderCtrl is able to analyze up to 38 micro facial gestures of travelers while they answer questions from airport border agents. Thanks to this, it can induce, with a reliability of 76%, whether people are telling the truth or lying, thus helping to detect suspicious behavior. In addition, it scans biometric data such as fingerprints, veins or arteries in the palm of the hand to help verify the identity of people.
These tools put the technology at the service of truth. Neither machines nor people are foolproof, but the combination of human intuition, based on professional experience, with the implementation of artificial intelligence solutions in law enforcement can help build a safer world for all.