Integrating user demographic parameters for mouse behavioral biometric-based assessment fraud detection in online education platforms

Online education systems have gained immense popularity due to their ubiquity, flexibility, openness, and accessibility. This has led many higher education institutions to incorporate online courses as part of blended or fully online learning. However, online assessment fraud remains a critical challenge. Conventional assessment fraud detection methods are often one-time, non-repudiable, invasive, expensive, and susceptible to spoofing. Even some advanced systems based on behavioral biometrics report comparatively lower accuracy, underscoring the ongoing challenge of achieving reliable user authentication. Furthermore, few research studies focus on behavioral biometric-based assessment fraud detection in online education platforms. To address these gaps, we introduce the UserID.AGE.GEN framework, which implements a cross-referencing fusion algorithm that integrates user demographic parameters, including age and gender, with mouse behavioral biometrics for user identity verification for online assessment fraud. Additionally, we collect novel task-specific data for our evaluation. Experimental results demonstrate that our method achieves promising results compared to some existing models, highlighting its strong performance and promising potential for broader application and future enhancement. A notable limitation of the proposed model is that it has not yet been evaluated using significantly larger external datasets, which may affect the generalizability of the results. Our evaluation was conducted using internally collected datasets. Additionally, the model has not been tested in real-world settings such as online education platforms, which may limit insights into its practical deployment.