Biometrics-based Password Identity Verification : Some Practical Issues and Solutions

Password is the most widely used identity verification method in computer security domain. However, due to its simplicity, it is vulnerable to imposter attacks. One way to add a shield to password is to employ keystroke dynamics, which is a biometric-based approach. Without manual preprocessing to remove noises and outliers resulted from typing inconsistencies, a poor detection accuracy results. Regarding this, we first introduce our previous research on automatic feature subset selection process, i.e. GA-SVM wrapper approach, which can automatically selects a relevant subset of features and ignores the rest, thus producing a better accuracy. Furthermore, a user authentication system often starts with an insufficient amount of data, which consequently guide the wrapper search to overfitting. In later stages of GA-SVM wrapper, diversified while semi-converged feature members are generated. Since the objective of feature selection exercise is to build a better ‘authenticator’, rather than to find out important features for data mining reasons, then the ensemble of these members can offer a good solution. In this article, we also propose an ensemble creation method, whose members are from GA-SVM feature selection procedure, to improve keystroke dynamics identity verification capability. Experimental results show that the proposed method is promising in dealing with such practical issues in keystroke dynamics identity verification problems.