A Framework of Bimodal Biometrics for E-assessment Authentication Systems

The global use of the internet has improved the growth of the educational sector over the years, while electronic assessments have turn out to be one of the major tools in the development of both non-academic and academic establishments. The effective assessment of a student is mostly perceived as one of the foremost challenges that is frequently experienced during online examination in that it can be very difficult to provide accurate user authentication. The requirement to secure and authenticate a user during e-assessments owing to the high rate of misconduct has led to the proposal of this research. The purpose is to examine potential threats to student authentication during e-assessments and propose a framework which uses a bi-modal authentication approach to provide successful authentication during e-assessment. In implementing this approach, we propose a framework that provides security to improve e-assessments by introducing authentication classifiers to demonstrate its application in biometrics technologies. The proposed model was evaluated based on set of thresholds using Accuracy, FAR and FRR as performance metrics. the proposed model gave a high accuracy of 94.52%. The single-modal model of keystrokes had percentage accuracy of 92.025% and face had percentage accuracy of 92.58%. This implies that the bimodal model integrating keystrokes and face outperforms the single-modal model of keystrokes and single-modal model of face respectively. The study concludes that the proposed model contributes to existing works on e-assessment systems by integrating keystrokes and face bimodal biometric to optimally minimize fraud and impersonation thereby providing accurate authentication of a user.