AI-powered tutoring system for automatic notes summarization and quiz generation

Salahuddin, Lizawati and Nor Afian, Irfan Syafie and Idris, Ariff and Abdul Rahim, Fiza (2025) AI-powered tutoring system for automatic notes summarization and quiz generation. International Journal of Research and Innovation in Social Science (IJRISS), IX (X). pp. 1156-1164. ISSN 2454-6186

Text 018510912202518432659.pdf Available under License Creative Commons Attribution. Download (358kB)

Abstract

The AI-Powered Tutoring System is developed to address the persistent challenges of limited personalized learning support and cognitive overload that students experience when managing and revising extensive academic materials. Despite the growing use of digital learning tools, few systems effectively integrate automated summarization and quiz generation to promote active, self-directed learning. This study bridges the gap by designing an intelligent tutoring platform that enhances comprehension and engagement through AI-driven automation. The system allows students to upload academic materials such as PDFs, text files, and PowerPoint presentations, which are then processed using DeepSeek’s natural language processing API to generate concise summaries and structured quizzes. The system was developed using a full-stack web architecture comprising React.js for the frontend, Node.js and Express for the backend, and PostgreSQL as the database. Guided by the Agile methodology, the development process was structured into iterative sprints encompassing key phases such as planning, design, development, and testing. The system integrates the DeepSeek API for natural language processing, enabling the platform to provide summarized lecture notes and structured quizzes tailored to each uploaded document. It also includes user authentication, file handling, progress tracking, and a personalized library, allowing users to manage their learning resources effectively. A user acceptance test (UAT) with 31 undergraduate students was conducted using a five-point Likert scale questionnaire. The UAT results show that perceived usefulness, perceived ease of use, capability, trustworthiness, attitude toward the system, and behavioral intention to use had a high acceptance rate. Theoretically, this work contributes to the advancement of AI-driven educational technology by integrating principles from self-directed learning and technology acceptance models. Pedagogically, it offers an innovative approach to improving accessibility, comprehension, and learner autonomy through adaptive AI tools for content summarization and assessment.

Item Type:	Article
Uncontrolled Keywords:	Artificial intelligence, Educational technology, DeepSeek, automated notes summarization, Quiz generation
Divisions:	Faculty of Information and Communication Technology
Depositing User:	Norfaradilla Idayu Ab. Ghafar
Date Deposited:	15 Jul 2026 00:05
Last Modified:	15 Jul 2026 00:05
URI:	http://eprints.utem.edu.my/id/eprint/29912
Statistic Details:	View Download Statistic

Actions (login required)

View Item