Randomized Point Selection in Graphical Password Authentication: A Secure Approach for Educational Systems

Abstract

The rise in cyber threats has really underscored the importance of having secure yet user-friendly ways to authenticate ourselves online. This study presents an innovative graphical authentication system that uses randomized point selection, which we’ve named the Encrypted Cued Click Point System (ECCPS). With this system, users can upload up to five personal images, each featuring three randomly generated points. When it comes time to authenticate, users need to pick one correct point from each image, creating a unique combination for secure access. By randomizing the point selection, ECCPS helps reduce the risks of shoulder-surfing and guessing attacks. We implemented the system on an academic result management platform and put it to the test with 50 students. We looked at various metrics, including how often users successfully authenticated, error rates, the time it took to authenticate, and overall user satisfaction to evaluate both usability and security. The results show that ECCPS strikes a great balance between strong security and user-friendliness, with over 90% of participants managing to authenticate within three tries and an average authentication time of 20 seconds. Participants provided positive feedback, pointing out the system's potential for wider use in handling sensitive information. This paper dives into the design, implementation, and evaluation of ECCPS, highlighting how it contributes to the evolution of graphical password systems.