AssessMate: Revolutionizing Assessment Design with AI
Yiliu Pan, Mingmei Zhang, Tzu-Yun Huang, Luojia Chen, Kanghua Qiu · 2025
In the era of Generative Artificial Intelligence (GenAI), traditional assessment methods, such as written tasks and multiple-choice questions, are increasingly susceptible to AI-generated responses, raising concerns about academic integrity. While AI detection tools have been proposed, they remain limited and often biased. Instead of focusing on detection, there is a growing shift toward redesigning assessments to integrate AI constructively. Learning Engineering (LE) provides a structured, iterative approach to designing, implementing, and evaluating technology-enhanced educational solutions. Our study applies an LE framework to the development of AssessMate, an AI-driven system that automates assessment design while ensuring alignment with educational objectives. By embedding principles of human-centered design, systematic evaluation, and pedagogical alignment, AssessMate supports educators in creating more authentic and robust assessments. This paper discusses the co-design process, user evaluations, and future refinements to enhance the system's effectiveness within the Learning Engineering paradigm.