An Adaptive Biometric Authentication System for Online Learning Environments Across Multiple Devices

Abstract

Online learning environments have become a crucial means to provide flexible and personalised pedagogical material, and a major driving cause is due to the COVID-19 pandemic. This has rapidly forced the migration and implementation of online education strategies across the world. Online learning environments have a requirement for high trust and confidence in establishing a student’s identity and the authenticity of their work, and this need to lessen academic malpractices due to increased online delivery and assure the quality in education has accelerated. In addition to this, due to the ubiquity of mobile devices such as smartphones, tablets and laptops, students use a variety of devices to access online learning environments. Therefore, authentication systems for online learning environments should operate effectively on those devices to authenticate and invigilate online students. Confidence in authentication systems is also crucial to detect cheating and plagiarism for online education as strong authorisation and protection mechanisms for sensitive information and services are bypassed if authentication confidence is low. In this paper, we examine issues of existing authentication solutions for online learning environments and propose a design for an adaptive biometric authentication system for online learning environments that will automatically detect and adapt to changes in the operating environment. Multi-modal biometrics are applied in the proposed system which will dynamically select combinations of biometrics depending on a user’s authenticating device. The adaptation strategy updates two thresholds (decision and adaptation) as well as the user’s biometric template they are using the authentication system.