A multimodal approach to mitigate cheating in online assessments

Cheating in online assessments remains a persistent challenge despite the presence of numerous research efforts and commercial solutions. In this context, we propose a comprehensive multimodal solution that not only detects cheating but places a primary emphasis on mitigation strategies, recognizing their efficiency. Our solution encompasses various modules, each tailored to combat specific forms of misconduct, and we provide a detailed illustration of the behaviors it targets. Our objective is to develop an effective and adaptable solution. By prioritizing mitigation, we aim to proactively prevent cheating. Furthermore, our system is designed with flexibility in mind, enabling adjustments to address emerging methods of cheating and evolving abnormal behaviors employed by students in digital assessments.