User Authentication Based on Keystroke Dynamics
Keystroke dynamics is a special authentication technique which identifies a person based upon his/her behavioural characteristics in order to avoid the shortcomings of the existing password-based system and to enhance the security.
Keystroke dynamics based biometric system does not require an extra hardware makes it cost effective and convenient. It is non-intrusive and non-obstructive making it a perfect choice for authentication. When a user types the password to get access of system, his/her keystrokes are collected. The latencies between successive keystrokes, keystroke durations, can be used to construct a unique profile for that individual. The authentic user of the computer is identified based on previously collected information of keystroke dynamics profiles. These profiles get continuously updated over a period of time in order to account for temporal changes in user behavioural biometric. The following figures depicts the pre-processing, training and authentication phase :
Using a small training sample, accuracies of 90% and 99% are achieved for the data-set of 220 login strings per user (40 strings from legal user + 180 strings from nine intruders) using Gaussian mixture model and two-layer feed-forward neural network, respectively, as classifier.
Patent:
1. Sudipta Mukhopadhyay , Rahul Arora, "A Keystroke dynamics based system for synthesis of imposter data and normalization of energy level for user authentication and method thereof", Patent Application: 837/KOL/2014 dated 08-Aug-2014 (Full specification filed in Aug-2015).
Publication:
1. Rajat Kumar Das, Sudipta Mukhopadhyay and Puranjoy Bhattacharya, "User Authentication Based on Keystroke Dynamics", IETE Journal Of Research, pages 229-239, Volume 60, Issue 3, 2014.
2. Rajat Kumar Das, Sudipta Muhkhopadhyay and Puranjoy Bhattacharya "Continuous multimodal biometric authentication for PC and handheld devices", IETE Journal of Education, pp 59-69 vol. 52, Issue 2, July-December 2011.