Data driven user authentication with multi-level behavior profiling
Date
2021-08
Authors
Fu, Shen
Major Professor
Advisor
Qiao, Daji
Guan, Yong
Duwe, Henry
Smiley, Ann
Wymore, Mathew L
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Abstract
The restriction of access to software systems is more important than ever. For example, critical data is increasingly being stored on web services that are accessible from anywhere in the world. Yet most primary authentication methods are still largely based on passwords, which are vulnerable to various attacks such as phishing scams and keyloggers. In this work, we are seeking innovative data-driven user authentication schemes for general software systems that can take the advantage of other factors beyond passwords. We dive into the study of user behavior at two levels: software-level behavior and mouse operation behavior. For software-level behavior, we propose a generic continuous authentication scheme, which supplements existing authentication schemes and works as an auxiliary layer to provide additional protection against impostors. For mouse operation behavior, we first conduct an in-depth study of mouse behavior based user authentication, and provide an overview of the existing works in different authentication scenarios and categories of schemes. Then, we propose a CNN-RNN combined neural network model for mouse behavior based user authentication, which takes raw sequential mouse data as input rather than relies on heuristic feature extraction. Furthermore, we propose a novel framework of hybrid human-machine learning, in which the user achieves a level of artificially-induced expertise to interact with a customized machine and therefore become easier to be recognized by machine learning algorithms. We implement this concept in a mouse-based user authentication system by introducing an angle offset to the standard mouse. The work presented here is intended to provide authentication schemes via other factors beyond passwords and therefore enhance the security of contemporary software systems.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
dissertation