| Title |
An Implementation of User Authentication System using Keystroke Dynamics-based and Adaptive Risk-based Authentication |
| Authors |
신지안(Ji-Ahn Shin) ; 이상협(Sang-Hyeop Lee) ; 이승현(Seung-Hyeon Lee) ; 이일현(Il-Hyn Lee) ; 이광재(Kwangjae Lee) |
| DOI |
https://doi.org/10.5370/KIEEP.2020.69.4.267 |
| Keywords |
User authentication; Adaptive risk-based authentication; Keystroke dynamics-based authentication; Machine learning |
| Abstract |
In this paper, we propose a system that has the convenience of single-factor authentication and security of multi-factor authentication at the same time. This system consists of keystroke dynamics-based authentication(KDA), which uses biometrics, and adaptive risk-based authentication(A-RBA), which uses a user access information. KDA applied the One-Class SVM model, which is an unsupervised learning method that uses small dataset, and A-RBA applied the KNN model, which is supervised learning through a dataset created as a virtual scenario. As a result of the experiment, FAR and FRR of KDA are 0.21 and 0.04, respectively, and accuracy, recall, precision, F1 score, and CVAA of A-RBA are 97.5%, 98.5%, 98.3%, 98.1% and 99.2%, respectively. This result is higher performance than other machine learning models such as Logistic Regression and SVM. This proposed system not only provides user convenience, but will also be a good way to overcome the increasingly emerging hacking attacks. |