IOSR Journal of Computer Engineering (IOSR-JCE) e-ISSN: 2278-0661,p-ISSN: 2278-8727 PP 11-14 www.iosrjournals.org International Conference on Emerging Trends in Engineering & Management 11 |Page (ICETEM-2016) A Review of Behaviometric Techniques for User Authentication Princy Ann Thomas, Lakshmi Priya A, Dr. Suvanam Sasidhar Babu Dept. of CSE (Cyber Security) Sree Narayana Gurukulam College of Engineering Kadayiruppu, Kerala, India Dept. of CSE (Cyber Security) Sree Narayana Gurukulam College of Engineering Kadayiruppu, Kerala, India Dept. of CSE Sree Narayana Gurukulam College of Engineering Kadayiruppu, Kerala, India Abstract: Biometrics is an authentication mechanism classified into physical biometrics and behavioural biometrics. Physical biometrics uses the physical attributes of the individual to verify the claimed identity. Behavioural Biometrics or behaviometrics uses measurable behavioural patterns of device usage to verify the identity of the individual. Physical biometrics would require a special hardware device to capture the characteristics of the claimed identity for verification purposes but behaviometric does not require one in most cases. Keystroke dynamics and mouse dynamics have been researched most frequently for the purpose of static authentication, periodic authentication and continuous authentication of users. In this paper we make a comparative survey on the effectiveness of keystroke dynamics, Keystroke sound, mouse dynamics and mouse gesture dynamics individually as a defense-in-depth mechanism for authentication purposes. Keywords: Behavioural Biometrics, Keystroke Dynamics, Mouse Dynamics, Static Authentication, Continuous Authentication, Equal Error Rate (EER). I. INTRODUCTION User Authentication is the process by which a device or application is able to verify the identity of an individual user or claimed identity. This allows privilege assignment and authorization according to user requirements. In the authentication process the credentials provided by the user are compared to stored values. If a match occurs the process is completed and user is granted authorization for access. Authentication mechanisms are generally categorized into three types. Type 1 is based on something the user knows like passwords. Type 2 is based on something the user has as in the case of credit cards or ATM cards. Type 3 is based on something the user is or something that is unique about the user. Biometrics falls in the Type 3 category. Biometrics is further categorized into physical biometrics and behavioural biometrics. Physical Biometrics uses the physical characteristics of an individual for authentication like fingerprint, iris, facial, retina, hand geometry, voice and vein pattern. Behavioural Biometrics is also referred to as behaviometrics. Behaviometrics is based on the behavioural pattern of an individual when using specific devices. These patterns are unique to the individual and can be used to verify the identification of the user. The most commonly researched techniques for authentication using behaviometrics are keystroke dynamics [1], [6] and mouse dynamics [10], [11], [12] as they are less vulnerable to effects of physical environmental changes. Added benefits of using these methods are that they do not require any additional equipment to capture the behavioural pattern. This paper focuses on recent techniques using keystroke dynamics and mouse dynamics for authentication. We compare research on the use of keystroke dynamics [3], keystroke sound [7], mouse dynamics based on mouse clicks [12] and mouse gesture dynamics [13]. Current research reports the results in terms of Equal Error Rate (EER) or False Acceptance Rate (FAR) and False Rejection Rate (FRR). The remainder of this paper is organized as follows. In Section II, we present an introduction on the behavioral biometric techniques that are reviewed for authentication in this paper. In Section III, we present findings from most recent work done in behaviometric authentication. In Section IV, we make a comparative study on the different research work presented. In Section V, we give our conclusions. II. Behaviometric Methods Behavioural biometrics makes use of measurable behavior patterns to recognize the identity of an individual. Most of these methods work in two stages such as enrollment stage and identification stage. In both data capture is followed by feature extraction using different algorithms or methods. In the enrollment stage, after feature extraction the data is stored into a database or as a template of the user profile. In the case of identification stage instead of storing the extracted features it is matched with the stored template. If a match occurs the user is authenticated.