Keystroke Biometric Studies on Short and Long Input Including Numerics : Intruder Detection using Keystroke Biometrics on Short Numeric Input and Long Text, Browser, and Spreadsheet Input

A keystroke biometric system was enhanced to capture raw keystroke data directly from an individual's computer system using an open source key logger originally designed for software testing. The key logger runs in the background capturing keystrokes directly from the operating system requiring no additional capture software, text entry window, or edit box for input. This allows the freedom to generate unrestricted entry from any application installed on a system. Long input data were collected from 20 participants using spreadsheet, browser, and text applications. Subjects were mainly students at the University and free to type whatever they desired without using copy tasks in any of these experimental scenarios. Short input data simulating a ten digit passcode were also collected from 30 participants each entering the same digit sequence strictly from the numeric keypad section of the keyboard. Verification experiments using novel approaches regarding feature sets and classification methods were run on both the unrestricted long input samples and static short passcode input. ROC curves were generated for each experiment using Equal Error Rate as the essential performance measure.