Organisations are challenged to keep applications and networks secure in the limits of cost-security balance maintenance. Relying on only userID and userPassword to authenticate users is neither practical nor efficient. Traditional security measures like one time passwords, tokens, access cards, PINs or device signatures are expensive, hard to deploy and add an extra difficulty at the applications usage. As we accelerate in the 21st century, new challenges appear. Elaborated measures to stop the unauthorized access to computer resources and information are being developed. The paper presents one safeguard based on authenticated access to resources via recognising some unique patterns in the user's typing rhythm: keystroke recognition. The process of key typing and its rhythm can disclose individual patterns, which combined form the basis of the biometric technology known as keystroke dynamics. Its main purpose is to confirm the identity of the user, rather than uniquely identify it. Keystroke recognition is simple to implement because it supports mainly a software implementation. Due to that, the deployment of systems based on keystroke recognition is made in low-stakes, computer-centric applications such as content filtering or digital rights management where the password to download the info is bolstered with by keystroke dynamic verification to prevent the password sharing. We have developed a fast and reliable scheme for keystroke recognition. Code has been tested on Jeffrey D. Allen's Keystroke Dynamics Dataset. Index Terms: Matlab, source, code, Keystroke recognition, online fraud, computer access security, pattern recognition, identity thefts, biometric authentication, keystroke dynamics.
Requirements:
Matlab
Comments not found