基於自動問題生成與知識追蹤之適性化學習系統設計與應用:以 Python 程式設計學習為例;Design and Application of an Adaptive Learning System Based on Automatic Question Generation and Knowledge Tracking: A Case Study of Python Programming Learning

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Network Learning Technology > Electronic Thesis & Dissertation > Item 987654321/96369

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/96369

Title:	基於自動問題生成與知識追蹤之適性化學習系統設計與應用:以 Python 程式設計學習為例;Design and Application of an Adaptive Learning System Based on Automatic Question Generation and Knowledge Tracking: A Case Study of Python Programming Learning
Authors:	呂浚宏;Hung, Lu-Chun
Contributors:	網路學習科技研究所
Keywords:	適性化學習;自動問題生成;知識追蹤;間隔學習;ChatGPT;Adaptive Learning;Automatic Question Generation;Knowledge Tracing;Spaced Learning;ChatGPT
Date:	2024-12-24
Issue Date:	2025-04-09 18:21:33 (UTC+8)
Publisher:	國立中央大學
Abstract:	因應大數據時代下的人工智慧熱潮，程式設計與資料分析能力成為業界最欠缺的實作技能之一。在教學過程中發現，教師在培養學生產業實戰程式設計能力時面臨諸多挑戰，例如無法確定學生是否有效吸收課堂所學知識、針對不同程度學生設計測驗題耗時費工以及無法對學生知識缺陷提出適性化學習建議等問題。為解決這些教學現場的問題，本研究欲導入所研發之適性化間隔學習系統，利用適性化推題機制輔以知識追蹤技術，提供學生適合的學習內容，期望能提高學生的學習成效，並讓教師掌握學生的學習狀況。此系統將結合生成式人工智慧技術，例如自動摘要、自動出題和角色扮演等功能。其中，本研究讓生程式人工智慧模型 ChatGPT 嘗試扮演 Python 程式課老師，進行自動出題(Automatic Question Generation, AQG)，減輕教師在設計測驗題上的負擔。通過知識追蹤技術對每個問題進行知識點標記，根據學生的回答結果，自動調整後續問題的難度和內容，實現適性化學習。本研究結果顯示，適性化學習系統確實能夠有效減輕教師負擔，提升學生學習效果，並促進教師與學生之間的互動。由 ChatGPT 所自動生成題目也具有與教師生成之題目品質相似，證實了 ChatGPT 作為教師角色的可行性;且透過結合知識追蹤的適性化推題機制，不僅使學生學習更加高效，也讓教師能夠更精確地了解學生的學習狀況，進而進行針對性的教學調整，這對於生程式人工智慧與現代教育的結合發展具有重要意義。;In response to the surge of artificial intelligence in the era of big data, programming and data analysis skills have become some of the most lacking practical abilities in the industry. During the teaching process, it was found that teachers face numerous challenges in developing students′ practical programming skills for the industry, such as being unable to determine whether students effectively absorb the knowledge taught in class, the time-consuming task of designing test questions for students of different levels, and the inability to provide tailored learning recommendations for students′ knowledge deficiencies. To address these issues in the teaching field, this research intends to introduce a developed adaptive spaced learning system. This system uses an adaptive question-pushing mechanism supplemented by knowledge tracking technology to provide students with suitable learning content, aiming to improve students′ learning outcomes and help teachers understand students′ learning statuses. The system will integrate generative AI technologies, such as automatic summarization, automatic question generation, and role-playing functions. In particular, this research will have the generative AI model ChatGPT attempt to act as a Python programming course teacher, generating questions automatically (Automatic Question Generation, AQG) to reduce the burden on teachers in designing test questions. Through knowledge tracking technology, each question will be marked with knowledge points, and based on students′ answers, the difficulty and content of subsequent questions will be automatically adjusted to achieve adaptive learning. The results of this study demonstrate that the adaptive learning platform effectively reduces the burden on teachers, enhances student learning outcomes, and fosters interaction between teachers and students. The questions automatically generated by ChatGPT are of similar quality to those created by teachers, confirming the feasibility of using ChatGPT in a teaching role. By integrating a knowledge-tracking adaptive question recommendation mechanism, the system not only makes student learning more efficient but also allows teachers to accurately understand students′ learning progress. This enables targeted teaching adjustments, which is significant for the integration of generative artificial intelligence and modern education.
Appears in Collections:	[Graduate Institute of Network Learning Technology] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	13	View/Open

社群 sharing

Loading...