Long-Text Keystroke Biometric Applications over the Internet

Biometrics, the computer-based verification or identification of an individual, is becoming more and more essential due to the increasing demand for high-security systems. Accurate and effective keystroke biometric technology can help provide a major boost to the security of electronic commerce, and it can help curb identify theft. We focus here on two keystroke biometric applications that operate over the Internet and that require a text input of several hundred characters. The first application can help identify a perpetrator of inappropriate or fraudulent Internet activity, and the second can verify the identity of a computer user, such as a student taking an online exam over the Internet. A Java applet collects the raw keystroke data over the Internet. Feature measurements are then extracted from the raw data, and a pattern classification system, initially a simple nearest neighbor classifier, is trained to make the appropriate application-dependent decision. We present preliminary experimental results on the effectiveness of our system in these two applications of keystroke biometrics.