Towards More Trustworthy Online Examinations: Formal Modeling and Verification Using CPNs

Abstract

Online examinations are among the most effective ways to evaluate students' learning outcomes. The reliable design of online examination systems is crucial to ensuring the correctness of examination procedures. While existing approaches focus on isolated aspects (for example, state-level verification via UPPAAL or secure data transmission via cryptography), they lack the capability for comprehensive verification of the entire examination workflow. To address this gap, we use colored Petri nets (CPNs), which enable end-to-end modeling of concurrent processes and data-token interactions. In this paper, we develop a comprehensive CPN-based framework to model the entire online examination workflow. We formalize eight critical properties (e.g., correct total score) using ASK-CTL and verify them through state-space exploration and scenario simulation in CPN Tools. The verification results demonstrate that the model satisfies all eight properties, with no deadlocks or property violations found among 62,292 state-space nodes. This work provides a formal verification framework for designing trustworthy online examination systems, reducing risks of logical errors and security vulnerabilities in real-world implementations.