IJIET 2025 Vol.15(9): 1768-1776
doi: 10.18178/ijiet.2025.15.9.2378
doi: 10.18178/ijiet.2025.15.9.2378
Beyond One-Size-Fits-All: An AI-Driven Approach for Personalized Quizzes Using Clustering and ChatGPT
Liyuan Liu1,*, Meng Han2, Seyedamin Pouriyeh3, Yiyun Zhou4,
and Nasrin Dehbozorgi5
1. Department of Decision and System Sciences, Saint Joseph’s University, United States
2. Intelligent Fusion Research Center, Zhejiang University, China
3. Department of Information Technology, Kennesaw State University, United States
4. The National Board of Medical Examiners, United States
5. Department of Software Engineering, Kennesaw State University, United States
Email: lliu@sju.edu (L.L.); mhan@zju.edu.cn (M.H.); spouriye@kennesaw.edu (S.P.); yyzhou@nbme.org (Y.Z.); dnasrin@kennesaw.edu (N.D.)
*Corresponding author
2. Intelligent Fusion Research Center, Zhejiang University, China
3. Department of Information Technology, Kennesaw State University, United States
4. The National Board of Medical Examiners, United States
5. Department of Software Engineering, Kennesaw State University, United States
Email: lliu@sju.edu (L.L.); mhan@zju.edu.cn (M.H.); spouriye@kennesaw.edu (S.P.); yyzhou@nbme.org (Y.Z.); dnasrin@kennesaw.edu (N.D.)
*Corresponding author
Manuscript received March 11, 2025; revised April 27, 2025; accepted May 23, 2025; published September 8, 2025
Abstract—With rapid advancements in Generative AI (GenAI), educators have the opportunity to enhance student engagement and learning through personalized quizzes. Despite their potential, the adoption of customized learning assessments remains limited due to challenges in student grouping, difficulty calibration, and content fairness. This study proposes a structured, three-step framework leveraging AI to address these issues. Firstly, diverse student data—including academic performance, behavior, interaction with learning materials, demographics, psychological attributes, and feedback—is aggregated and normalized into multi-dimensional vectors. Kmeans clustering with Euclidean distance is then applied. Secondly, detailed profiles are created for each cluster by calculating their centroid, reflecting the unique characteristics and preferences of students. Finally, these profiles guide a GenAI system-ChatGPT to generate personalized quiz questions relevant to each group’s learning style and field of study. Implementing this approach with 105 business statistics students at a university in the USA, statistical tests demonstrated significant improvement in student performance on customized quizzes compared to traditional assessments. The findings underscore the transformative potential of AI-driven personalization in educational settings, promoting more effective, tailored learning experiences.
Keywords—artificial intelligence, generative artificial intelligence, customized learning, students clustering
Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).
Keywords—artificial intelligence, generative artificial intelligence, customized learning, students clustering
Cite: Liyuan Liu, Meng Han, Seyedamin Pouriyeh, Yiyun Zhou, and Nasrin Dehbozorgi, "Beyond One-Size-Fits-All: An AI-Driven Approach for Personalized Quizzes Using Clustering and ChatGPT," International Journal of Information and Education Technology, vol. 15, no. 9, pp. 1768-1776, 2025.
Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).