User Identity Authentication and Identification Based on Multi-Factor Behavior Features
- Resource Type
- Conference
- Authors
- Ding, Xing; Peng, Changgen; Ding, Hongfa; Wang, Maoni; Yang, Hui; Yu, Qinyong
- Source
- 2019 IEEE Globecom Workshops (GC Wkshps) Globecom Workshops (GC Wkshps), 2019 IEEE. :1-6 Dec, 2019
- Subject
- Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Engineering Profession
General Topics for Engineers
Power, Energy and Industry Applications
Signal Processing and Analysis
Authentication
Mice
IP networks
Feature extraction
Support vector machines
Hidden Markov models
Presses
- Language
Behavior-based identity authentication has been of research interests due to its low cost and the fact that such authentication factors cannot easily copied nor stolen. However, single behavior feature authentication usuall cannot gaurantee sufficient accuracy. In this paper, a multi-factor behavior features identity authentication method is proposed by combining factors of mouse, keystroke and address. We apply support vector machine by using the information divergence of features as the weight of the linear kernel function to solve quantification's problem of features influence degree on authentication. The tailored support vector machine is applied to multi-factor behavior features to identity authentication. Our experimental results show that the improved support vector machine is better than both the linear and Gaussian kernel support vector machines for identity authentication in accuracy. The proppsed method can be applied to large volume users' identity authentication cases where their behavior features are likely to be similar.