Two-Factor Authentication Application Using Artificial Intelligence to Support Academic Information Systems

Integrating facial recognition technology is crucial for enhancing the security of online learning and consultation platforms, particularly in computer technology and engineering education. This project aims to synthesize components and technologies to develop a prototype of a two-factor authentication system that leverages artificial intelligence technology. The goal is to enhance the efficiency and capacity of identification verification. Moreover, a comprehensive evaluation will be carried out to determine the prototype's performance and level of acceptance. A purposive random sampling strategy was employed to select 40 participants from the Computer Technology and Digital Program at Rajabhat Maha Sarakham University, including students and faculty members. The research instruments comprised a thorough questionnaire encompassing several facets of the academic information system's development, a prototype of a two-factor authentication system, a quality assessment form, and a questionnaire to gauge the information system's acceptance. The performance criteria comprised accuracy, precision, recall, f1-score, average, and standard deviation. The findings indicated that the authorized prototype comprised four distinct modules: an authentication module, a member module, an information module, and a management module. The evaluation results for identifying faces, using CNN Face Detector, VGG-Face net, and classification by Logistic Regression, attained a remarkable accuracy of 83.54%. The precision and recall values were 0.84 and 0.88, respectively. The evaluation findings indicate that the overall quality is quite acceptable, with a mean score of 4.44 and a standard deviation of 0.55. Similarly, the user satisfaction with the 2FA system prototype is high, as indicated by a mean score of 4.54 and a standard deviation of 0.51.