Mathematicians in Kursk claim to have developed typing style identification software
The proposed IT product is even capable of identifying the bank client by the way he or she enters the plastic card’s pin code
KURSK, December 17. /TASS/. Scientists at the Kursk State University say they have developed software capable of identifying the author of the message being keyed in by the typing style, which is as unique as a fingerprint. As the chief of the research project, lecturer at the mathematical analysis and applied mathematics department, Leonid Kryzhevich, has told TASS, he sees potential uses of such software in education, at tight security facilities, and even in ATMs. The proposed IT product is capable of identifying the bank client by the way he or she enters the plastic card’s pin code.
“When someone types a text, some 40 muscles are at work simultaneously, and each individual uses them differently, with varying degrees of strength and intensity. Apart from the time spent on finding the keys, which is a very important parameter but depends directly on the individual’s ability to learn, the time during which this or that key remains pressed is very important, too. It lasts several microseconds, and each such tap on a key or button is unique,” Kryzhevich explained.
The scientist claims that the team he leads at the Kursk State University has developed a probability theory and statistics-based software capable of identifying the computer’s user with a 96-percent degree of certainty.
The software has two main functions — memorizing the users’ typing style and identifying it during each following typing session, which may prove very useful for control of access to sensitive facilities. If the program decides that the pin code or password has been keyed in by a stranger, access is denied.
Kryzhevich argues that the analysis of four to six taps on buttons or keys is enough for the program to make a decision. He speculates that its pioneer version will hit the market next year.