生成式AI賦能的回饋機器人用於改善學習成效;Improving Learning Achievement with a Generative AI-Enabled Feedback Chatbot

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/95439

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/95439

Title:	生成式AI賦能的回饋機器人用於改善學習成效;Improving Learning Achievement with a Generative AI-Enabled Feedback Chatbot
Authors:	蘇聖益;Su, Sheng-Yi
Contributors:	資訊工程學系
Keywords:	生成式AI;機器人;概念回饋;程式設計回饋;學習成效;generative AI;robot;concept feedback;programming feedback;learning achievement
Date:	2024-07-10
Issue Date:	2024-10-09 16:51:36 (UTC+8)
Publisher:	國立中央大學
Abstract:	本研究介紹了PyFeedbacker，一款結合了先進的生成式AI技術和機器人功能的教育工具。PyFeedbacker的核心特色在於其能夠提供針對性的概念回饋與程式設計回饋，概念回饋透過生成式AI比對範例總結與學生總結，提供關鍵字與概念提示，幫助學生進一步完善其概念，並生成針對性回饋，從而精確提供學生所需的概念回饋；程式設計回饋能回答學生在程式設計實作過程中的提問， PyFeedbacker 能分析學生的程式碼，並提供錯誤診斷與改進建議作為程式設計回饋，從而使學生在程式設計的學習過程中能更快地識別和修正問題，進而提高解決問題的能力。本研究展示了生成式AI技術在教育應用上的可能性，有望在教育回饋產生上減少對於教師、助教的依賴，進而改變教育資源的配置方式。本研究詳細闡述了PyFeedbacker的系統架構設計、生成式AI技術的整合過程，以及其在教育實踐中的實驗設計及成果。在實證研究方面，我們在大學的Python 程式設計課程中對 PyFeedbacker 進行了實驗，以評估其在提高學生學習成效和激發學習動機、學習策略方面的效果。實驗結果顯示，學生在使用PyFeedbacker進行學習時，不僅學習成效得到提升，學習動機與策略也隨之增強。本研究不僅證實了PyFeedbacker 在教育領域的應用價值，也展示了生成式AI技術與教育整合的潛力及生成式AI-人類協作模式的實際應用。;This study introduces PyFeedbacker, an educational tool that combines advanced generative AI technology with robotic functions. The core feature of PyFeedbacker is its ability to provide targeted conceptual feedback and programming feedback. Conceptual feedback offers hints related to concepts and keywords to help students improve their understanding. This feature uses generative AI to compare examples with student summaries and generates corresponding feedback. Programming feedback addresses students′ questions during programming exercises. PyFeedbacker analyzes students′ code, provides error diagnosis, and offers improvement suggestions, enabling students to quickly identify and correct issues, thus enhancing their problem-solving skills. The study demonstrates the potential of generative AI technology in educational applications, promising to reduce reliance on teachers and teaching assistants for feedback, and potentially changing the allocation of educational resources. The research details the system architecture design of PyFeedbacker, the integration process of generative AI technology, and the experimental design and outcomes in educational practice. In the empirical research, we conducted experiments with PyFeedbacker in a university Python programming course to evaluate its effectiveness in improving students′ learning outcomes and motivating learning strategies. The results showed that students not only improved their learning outcomes but also enhanced their motivation and strategies when using PyFeedbacker. This study not only confirms the value of PyFeedbacker in the field of education but also showcases the potential of integrating generative AI technology with education and the practical application of the generative AI-human collaboration model.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	15	View/Open

社群 sharing

Loading...