A methodology for identifying internet users by their typing style—so-called “keystroke dynamics”—is being developed in Russia. The project is being carried out by the Ural State Law University with funding from a grant provided by the Russian Science Foundation, the university’s press service told Uralinformburo. The developers say the goal is to create a “legally significant forensic methodology” for law-enforcement agencies that would make it possible to identify the author of a text when traditional digital data—such as an IP address or an account—cannot be used. The method is primarily intended for investigating cases involving the dissemination of extremist materials, “fake news,” as well as identifying the authors of anonymous threats and creators of forged documents, the university added.
According to the developer, a person’s “keystroke dynamics” are unique, “like fingerprints.” The proposed method is based on analyzing dozens of biometric and behavioral parameters, including typing rhythm, key hold time, intervals between keystrokes, patterns of correcting mistakes, typical typos, and even the choice between the spellings “e” and “yo.” The developer says these characteristics persist even when a previously typed text is edited. To collect the data, a special program records every keystroke with millisecond precision. The information is then processed using specific formulas to create a user’s “digital portrait.”
More than 20 volunteers were recruited to test the algorithms. As part of the experiment, participants were given a laptop with the special software and asked both to retype set texts and to write essays on assigned topics in order to capture differences in typing behavior, junior researcher Anna Sosnovikova said.
The developers plan to test the methodology under real investigative conditions and then transfer it to law-enforcement agencies for practical use. The project is led by Dmitry Voronkov, head of the Gerasimov Department of Criminalistics.
Software for recording keyboard typing parameters had previously been developed by researchers at the Russian Technological University (RTU MIREA). It was proposed for use in corporate information systems to protect against unauthorized access.
