User Authentication Based on Keystroke Dynamics

ABSTRACT

This paper presents a technique to verify user identity using keystroke dynamics from short text, namely the computer login string. The keystroke behavioural pattern is obtained when a person types with a QWERTY keyboard. Two features hold time of an individual key and the latency of the consecutive keystrokes is used for authentication. 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. The paper then proceeds to a comprehensive study to explain how the accuracy varies with the length of the input string and with negative data in the training set.

Keywords:

• Behavioural biometric
• Continuous authentication
• Free text
• Gaussian mixture model (GMM)
• Hold time
• Latency time
• Keystroke dynamics
• Neural network (NN)

ACKNOWLEDGEMENTS

This work has been supported by Signal Processing and Machine Learning in Pervasive Healthcare project sponsored by Intel Technologies India Pvt. Ltd, Bangalore. We would like to thank Dr Satish Prasad Rath, Sreenivas Subramoney, and Sumeet Verma of Intel Technologies India Pvt. Ltd. for their constant encouragement and support. We are also thankful to co-workers Lavanya Sainik, Subhadeep Mukhopadhyay, and our laboratory in-charge, Mr Arumoy Mukhopadhyay, for their helpful suggestions and advice at the moments of crisis.

Additional information

Notes on contributors

Rajat Kumar Das

Rajat Kumar Das received his B.E. degree in electronics and communication engineering from Visvesvaraya Technological University, Karnataka, India, in 2005. He worked as an engineer in Cranes software international limited, Bangalore, India, for 4 years. At present he is an MS student in IIT-Kharagpur, under Prof. Sudipta Mukhopadhyay in the Electronics & Electrical Communication Engineering, where he is working on biometric recognition.

Sudipta Mukhopadhyay

Sudipta Mukhopadhyay received his B.E. (Electrical) from Jadavpur University, India, in 1988, M.Tech. and Ph.D. in Electrical Engineering from the Indian Institute of Technology, Kanpur in 1991 and 1996, respectively. After 10 years in various industries – Tata Consultancy Services Limited, Satyam Computers, Silicon Automation Systems (SASKEN Communications), General Electric India Technology Center, and Philips – he joined the Indian Institute of Technology, Kharagpur in 2005, as an assistant professor. At present he is serving IIT Kharagpur as an associate professor.

Puranjoy Bhattacharya

Puranjoy Bhattacharya received his B.Tech., M.Tech., and Ph.D. in Electrical Engineering from the Indian Institute of Technology, Kanpur, India, in 1990, 1993 and 1999, respectively. Since 1998, he has been working in industrial R&D in the areas of signal processing, pattern recognition and machine intelligence, particularly as applied to biomedical, audio and speech signals. He currently works at Intel Labs at Bengaluru, India.