Digraph Approximation with an Adaptation Technique for Mobile User Authentication through Keystroke Dynamics
Christy James Jose1, Jijo Francis2, Rajasree M.S3
1Christy James Jose, Department of Electronics and Communication, Govt. Engg. College Barton Hill, Thiruvananthapuram (Kerala), India.
2Jijo Francis, Department of Electronics and Communication, Govt. Engg. College Barton Hill, Thiruvananthapuram (Kerala), India.
3Rajasree M.S, Director and Professor of Indian Institute of Information Technology and Management – Kerala, Thiruvananthapuram (Kerala), India.
Manuscript received on 15 October 2015 | Revised Manuscript received on 25 October 2015 | Manuscript Published on 30 October 2015 | PP: 45-51 | Volume-5 Issue-1, October 2015 | Retrieval Number: A4280105115/15©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Mobile devices have evolved at a proliferating rate and are now used in almost all aspects of life. With these the ability to store potentially private or sensitive information on these devices has also increased. Hence an intrusion detection and prevention system is a necessity for preserving the confidentiality and integrity of users. Keystroke dynamics which refers to detailed typing pattern of a person is used to model user behavior and use the so formed footprint for user identification and intrusion detection. A neural network based system using monograph and digraph timings with digraph approximation and adaptation technique is proposed for keystroke dynamics in mobile devices for free text data. With adaptation mechanism, the missing monographs and digraphs and also the time bound variations of user keystroke time variations are captured and adapted. The combined use of digraph approximation and adaptation yields a False Acceptance Rate (FAR) and False Rejection Rate (FRR) of 0% for 22 users. The impact of adaptation on other performance measures like accuracy, specificity, sensitivity and Mean Square Error(MSE) is also studied.
Keywords: Keystroke Dynamics, Intrusion Detection, Adaptation Mechanism, Keystroke Authentication.
Scope of the Article: Mechanical Design