Authentication System Based on Keystroke Dynamics
Author
Zaharia, Sorin
Term
4. term
Publication year
2018
Submitted on
2018-06-07
Pages
103
Abstract
Virksomheder og enkeltpersoner er afhængige af at holde information sikker. Fordi nye angreb på it-systemer opstår hurtigt, har vi brug for nye teknikker til at verificere, hvem der bruger et system. Dette projekt introducerer tastetryksdynamik (en autentifikationsmetode, der genkender personer på måden, de skriver, fx deres rytme og timing) og foreslår et system, der kan kontrollere en brugers identitet løbende, ikke kun ved login. I modsætning til engangslogins vurderer systemet frit indtastet tekst, mens brugeren arbejder, og sammenligner den med brugerens typiske skriveadfærd. Projektet beskriver flere metoder til at udtrække kendetegn fra tastetryk, som omsætter timing til mønstre for at fange den enkeltes skriveadfærd.
Companies and individuals rely on keeping information safe. Because new attacks on computer systems appear quickly, we need new techniques to verify who is using a system. This project introduces keystroke dynamics (an authentication method that recognizes people by how they type, such as their rhythm and timing) and proposes a system that can check a user's identity continuously, not just at log-in. Unlike one-time logins, the system evaluates freely typed text as the user works and compares it to the user's typical typing pattern. The project describes several feature extraction methods that turn keystroke timing into patterns to capture an individual's typing behavior.
[This abstract was generated with the help of AI]
Keywords
Documents