Scholarlink

The Rapid development of online learning has suggested the pressing need for more dynamic and personalized testing processes. Conventional e-learning testing is based on static tests and feedback delayed in time that cannot adapt to the dynamic and heterogeneous demands of learners. The article explores the potential of Artificial Intelligence (AI)—and, more specifically, Generative AI and Computerized Adaptive Testing (CAT)—to enhance online testing through real-time individualized feedback and responsiveness to the performance of each student. With the use of AI techniques such as reinforcement learning, neuro-fuzzy systems, and course sequencing algorithms, the article discusses how adaptive testing and individualization of learning paths can serve formative and summative testing. The overview summarizes recent research and real-world implementations (e.g., Duolingo and Knewton) to demonstrate how AI renders content presentation adaptive, measurement of skills accurate, and learning more interactive. Certain key issues like algorithmic bias, data privacy, and academic integrity are touched upon as well.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing E-learning Evaluation Through Generative AI: Personalizing Assessment and Feedback During Learning Experience

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Taha El Ghadraoui,
  • Hiba Asri,
  • Zahi Jarir,
  • Abdelali Rochdi

摘要

The Rapid development of online learning has suggested the pressing need for more dynamic and personalized testing processes. Conventional e-learning testing is based on static tests and feedback delayed in time that cannot adapt to the dynamic and heterogeneous demands of learners. The article explores the potential of Artificial Intelligence (AI)—and, more specifically, Generative AI and Computerized Adaptive Testing (CAT)—to enhance online testing through real-time individualized feedback and responsiveness to the performance of each student. With the use of AI techniques such as reinforcement learning, neuro-fuzzy systems, and course sequencing algorithms, the article discusses how adaptive testing and individualization of learning paths can serve formative and summative testing. The overview summarizes recent research and real-world implementations (e.g., Duolingo and Knewton) to demonstrate how AI renders content presentation adaptive, measurement of skills accurate, and learning more interactive. Certain key issues like algorithmic bias, data privacy, and academic integrity are touched upon as well.