Mitigating insider threat by profiling users based on mouse usage pattern

Authentication is the process of proving who a user or program is when accessing a system. Three-factor authentication with something you know (password), something you are (biometric), and something you own (token) is the best way to prevent unauthorized access. One of these steps, biometric authentication, performs this verification by checking distinctive biological or behavioral characteristics. Behavioral biometrics is the field of study concerned with the measurement of uniquely descriptive and measurable patterns in human activity. Behaviors that can be used in this study area are processes that are subconsciously controlled by measurable patterns such as keyboard use, gait, mouse use. A well-perceived pattern allows us to constantly check the user's authentication. In this study, the detection of the mouse usage patterns of the users for the company and the inhibition of the insider attack vectors are studied. The studies on the described biometric model are carried out on the Balabit Mouse Dynamics and Bogazici Mouse Dynamics datasets. This area has been contributed by feature studies in the frequency domain, the use of machine learning methods based on use cases, and the procedures of fusing information from various detectors. We presented our findings as a security layer by designing how the proposed procedures could work as a whole as a service.