[ad_1]
Spelling correctors and grammar proofers have been around for decades, but what if software could also be used as a lie detector for textual documents? It may seem crazy, but it's a topic that is already being worked on by researchers at Cardiff University in the UK and the Spanish University Charles III in Madrid – and that seems to be paying off. What computer scientists have developed is a tool that could help police identify false statements about armed robbery. This is a serious problem in some parts of the world and can lead to a host of negative consequences – such as false claims for insurance, loss of police time, and possibly unjustified arrests. Filing a false police statement can result in heavy fines and even jail time.
The new text analysis tool, called VeriPol, uses machine learning techniques to examine information such as the type of items reported as stolen and the authors' descriptions. According to its creators, it can then identify fake police reports with a precision of more than 80%. Common recurrences in false reports include a greater focus on stolen property than the incident, a lack of specific details, limited details about the attacker, and lack of witnesses or solid evidence.
"The main goal is to create a tool that can be useful for early detection of these fake police reports, so that it can complement the expertise of the police," said Dr. Jose Camacho Collados, research associate at Cardiff University, Digital Trends. . "In particular, in [our research] we are specifically looking at whether such a tool can be automatically learned from the textual content of reports, written by the police, and applied to new police reports. We used classical machine learning methods by taking as input police reports known to be true or false. The main conclusions are that people are lying using similar models and that a classical machine learning approach can detect these models and apply this problem effectively with a high degree of accuracy. "
VeriPol has been used in several pilot studies, which have shown how it can be used successfully to filter false flight reports. "We have now integrated this system into the architecture of the Spanish police, as well as an integrated training program, and it is used in a large number of police stations in Spain," continued Camacho Collados. "As a side effect – and perhaps more importantly – we expect that raising awareness and communication of this problem will further prevent people from filing false police reports."
Source link
