應用大型語言模型於個人化學習：AI 輔助重點標記與 LLM 生成 Flashcards 於行動應用程式

簡易檢索 / 詳目顯示

回結果列表

研究生：	陳佑軒 Chen, Yu-Hsuan
論文名稱：	應用大型語言模型於個人化學習：AI 輔助重點標記與 LLM 生成 Flashcards 於行動應用程式 Applying Large Language Models to Personalized Learning: AI-Assisted Highlighting and LLM-Generated Flashcards in Mobile Applications
指導教授：	吳尚鴻 Wu, Shan-Hung
口試委員:	黃俊龍 Huang, Jiun-Long 張永儒 Chang, Yung-Ju 高宏宇 Kao, Hung-Yu
學位類別：	碩士 Master
系所名稱：	電機資訊學院 - 資訊系統與應用研究所 Institute of Information Systems and Applications
論文出版年：	2025
畢業學年度：	113
語文別：	中文
論文頁數：	54
中文關鍵詞：	大型語言模型、重點標示、抽認卡、行動學習應用
外文關鍵詞：	LLMs, Highlight, Flashcards, M-Learning
相關次數：	點閱：3 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

傳統學習方式在提供高效、個人化學習資源方面仍存在一定的局限性。當學生需要自行尋找或製作複習教材時，這些過程通常耗時且繁瑣，往往降低學習動力與效率。因此，如何運用技術自動生成多樣化且符合個人需求的學習資源，成為提升學習效果的重要課題。

近年來，大型語言模型（LLM）在學習技術領域的人機互動應用（Human-Computer Interaction in Learning Technologies）中展現了顯著的潛力。LLM擅於整合多種領域知識並挖掘其間的關聯性，使生成的學習資源不僅更具多樣性，亦更能貼近使用者的需求，從而有效提升學習效率與用戶體驗。

本研究的主要貢獻包括以下三點：
設計與實現：針對學習資源生成的個人化需求，開發了一款基於Android的應用程式，用於生成Flashcard這類以圖文形式簡化知識傳遞的學習工具。研究模擬人工手動生成對比經由 LLM 自動且優化生成的卡片內容，在學習成效及使用者體驗的影響。
實驗與分析：收集使用者在實驗應用程式的使用數據、考試結果與訪談內容，分析 LLM 替代傳統學習資源生成方式的有效性，並探討個人化調整策略如何在滿足人機互動設計理念的同時，實際提升學習成效。
拓展與展望：本研究聚焦於Flashcard作為學習工具的應用，探討其在資訊輔助學習中的成效的同時，也對為未來研究LLM在其他學習工具上的應用提供基礎支持。

本研究不僅驗證了LLM在學習資源個人化生成中的潛力，亦為結合人工智慧技術的創新學習模式提供了有力證據，進一步促進人機協作學習的發展。

Traditional learning methods still face limitations in providing efficient and personalized learning resources. When students need to create or search for review materials themselves, these processes are often time-consuming and cumbersome, which can reduce learning motivation and efficiency. Therefore, leveraging technology to automatically generate diverse and tailored learning resources has become a critical challenge in enhancing learning outcomes.

In recent years, large language models (LLMs) have demonstrated significant potential in the domain of human-computer interaction in learning technologies. LLMs excel at integrating knowledge from various fields and uncovering their interconnections, making the generated learning resources not only more diverse but also better aligned with user needs, thereby effectively improving learning efficiency and user experience.

The main contributions of this study are as follows:
Design and Implementation: To address personalized needs in learning resource generation, we developed an Android-based application for creating tools such as flashcards, which simplify knowledge transmission through text and visuals. The study compares manually created flashcards with those optimized and generated by LLMs, examining their impact on learning outcomes and user experience.
Experiment and Analysis: By collecting user data from the application and exam results, the study analyzes the effectiveness of replacing traditional learning resource generation methods with LLMs. It also explores how personalization strategies can enhance learning outcomes while adhering to human-computer interaction design principles.
Extension and Outlook: This research focuses on the application of flashcards as a learning tool, investigating their effectiveness in information-assisted learning. It also lays the foundation for future studies exploring the application of LLMs in other learning tools, such as mind maps.

This study not only validates the potential of LLMs in personalized learning resource generation but also provides strong evidence for innovative learning models that integrate AI technologies, further advancing human-computer collaborative learning development.

Abstract (Chinese) I
Abstract II
Contents IV
List of Figures VII
引言 1
相關研究 3
1 標註對學習成效的影響 . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 LLM 在生成個人化學習資源中的潛力 . . . . . . . . . . . . . . . . . 4
3 Prompt 設計的架構與調整 . . . . . . . . . . . . . . . . . . . . . . . . 4
4 Flashcard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5 現有學習應用程式的發展趨勢 . . . . . . . . . . . . . . . . . . . . . 5
主要方法 6
1 Prompt 設定 ＆ 卡片內容生成流程 . . . . . . . . . . . . . . . . . . . 8
2 假設與評估方式 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1 RQ1: LLM 輔助學習是否能提升學習動機與吸引力，使學
生更容易跨越學習動機門檻，從而增強學習意願？ . . . . . 8
2.2 RQ2: LLM 輔助生成的 Flashcard 結合使用者標記 ( High-
light ) 後，是否能更精確聚焦於學生關注的知識點，以提
升學習意願與成效？ . . . . . . . . . . . . . . . . . . . . . . 9
2.3 AI 輔助標記重點與 LLM 輔助生成 Flashcard 對使用者學習
體驗的影響是否存在相關性？ . . . . . . . . . . . . . . . . . 10
3 實驗目的 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4 實驗流程 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.1 受測者屬性與前測設計 . . . . . . . . . . . . . . . . . . . . . 11
5 應用程式功能 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.1 學習部分 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.2 卡片生成規則 . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.3 複習方式 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
實驗結果 23
1 使用數據 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.1 應用程式內數據 . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.2 應用程式外數據 . . . . . . . . . . . . . . . . . . . . . . . . . 27
2 測驗數據 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1 答題差距分析 . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2 學習前後測試結果 . . . . . . . . . . . . . . . . . . . . . . . 29
3 訪談數據 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.1 期望新增的功能 . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2 使用過程中遇到的問題 . . . . . . . . . . . . . . . . . . . . . 31
3.3 對現有功能的喜好比例 . . . . . . . . . . . . . . . . . . . . . 33
3.4 其他訪談部分 . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4 分析結論 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1 數據分析 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2 假設驗證 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
限制與未來研究 38
結論 39
A Historical Event Prompts 49
B English Vocabulary Prompts 52
                                

[1] Ali Husain Almadhoob, Akbar Sayed Kadhem Saleh, and Fatema Akbar. “QuizWiz: Integrating Generative Artificial Intelligence in an Online Study Tool”. In: Proceedings of the 2024 7th International Conference on Big Data and Education. ICBDE ’24. 2025, Pages 87 –96. DOI: 10.1145/3704289.3704296. URL: https://doi.org/10.1145/3704289.3704296.

[2] Hina Amin and Munawar Sultana Mirza. “Comparative study of knowledge and use of Bloom’s digital taxonomy by teachers and students in virtual and conventional universities”. In: Asian Association of Open Universities Journal. 2020,Pages 223–238. URL: https://www.emerald.com/insight/content/doi/10.1108/aaouj-01-2020-0005/full/html.

[3] AP United States History. URL: https://apstudents.collegeboard.org/courses/ap-united-states-history.

[4] David Baidoo-Anu and Leticia Owusu Ansah. “Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning”. In: SSRN 4337484. 2023, Volume: 7, Issue No: 1 Pages: 52–62. DOI: 10.61969/jai.1337500. URL: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337484.

[5] Yasmine Belghith et al. “Testing, Socializing, Exploring: Characterizing Middle Schoolers’ Approaches to and Conceptions of ChatGPT”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI ’24. 2024, Article No.: 276, Pages 1 –17. DOI 10.1145/3613904.3642332. URL: https://doi.org/10.1145/3613904.3642332.

[6] Kenneth E. Bell and John E. Limber. “Reading Skill, Textbook Marking, and Course Performance”. In: Literacy Research and Instruction. 2009, Pages 56–67. DOI: 10.1080/19388070802695879. URL: https://doi.org/10.1080/19388070802695879.

[7] Jesse Josua Benjamin et al. “Responding to Generative AI Technologies with Research-through-Design: The Ryelands AI Lab as an Exploratory Study”. In: Proceedings of the 2024 ACM Designing Interactive Systems Conference. DIS’24. 2024, Pages 1823 –1841. DOI: 10.1145/3643834.3660677. URL: https://dl.acm.org/doi/10.1145/3643834.3660677.

[8] Eason Chen et al. “GPTutor: Great Personalized Tutor with Large Language Models for Personalized Learning Content Generation”. In: Proceedings of the Eleventh ACM Conference on Learning @ Scale. L@S ’24. 2024, Pages 539 – 541. DOI: 10.1145/3657604.3664718. URL: https://dl.acm.org/doi/10.1145/3657604.3664718.

[9] Xiuge Chen et al. “Characteristics of Deep and Skim Reading on Smartphones vs. Desktop: A Comparative Study”. In: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. CHI ’23. 2023, Article No.: 516, Pages 1 –14. DOI: 10.1145/3544548.3581174. URL: https://doi.org/10.1145/3544548.3581174.

[10] E. Dickey, A. Bejarano, and C. Garg. Innovating Computer Programming Pedagogy: The AI-Lab Framework for Generative AI Adoption. 2023. DOI: 10.48550/arXiv.2308.12258. URL: https://arxiv.org/abs/2308.12258.

[11] Jacob Doughty et al. “A Comparative Study of AI-Generated (GPT-4) and Human-crafted MCQs in Programming Education”. In: Proceedings of the 26th Australasian Computing Education Conference. ACE ’24. 2024, pp. 114–123. DOI: 10.1145/3636243.3636256. URL: https://doi.org/10.1145/3636243.3636256.

[12] Fiona Draxler, Albrecht Schmidt, and Lewis L. Chuang. “Relevance, Effort, and Perceived Quality: Language Learners’ Experiences with AI-Generated Contextually Personalized Learning Material”. In: Proceedings of the 2023 ACM Designing Interactive Systems Conference. DIS ’23. 2023, Pages 2249 –2262. DOI: 10.1145/3563657.3596112. URL: https://doi.org/10.1145/3563657.3596112.

[13] Fiona Draxler et al. “Auto-Generating Multimedia Language Learning Material for Children with Off-the-Shelf AI”. In: Proceedings of Mensch und Computer 2022. MuC ’22. 2022, Pages 96 –105. DOI: 10.1145/3543758.3543777. URL: https://doi.org/10.1145/3543758.3543777.

[14] Duolingo. “Learn a language for free”. URL: https://www.duolingo.com/.

[15] ETS. The GRE® General Test. URL: https://www.ets.org/gre/test-takers/general-test/prepare/content/verbal-reasoning.html#accordion-9f58105fc6-item-88093eca37.

[16] M. Firat. How Chat GPT Can Transform Autodidactic Experiences and Open Education? 2023. DOI: 10.31219/osf.io/9ge8m. URL: https://doi.org/10.31219/osf.io/9ge8m.

[17] Andre Pimenta Freire, Paula Christina Figueira Cardoso, and Andre de Lima Salgado. “May We Consult ChatGPT in Our Human-Computer Interaction Written Exam? An Experience Report After a Professor Answered Yes”. In: Proceedings of the XXII Brazilian Symposium on Human Factors in Computing Systems. IHC’23. 2024, Article No.: 6, Pages 1 –11. DOI: 10.1145/3638067.3638100. URL: https://doi.org/10.1145/3638067.3638100.

[18] Yuichiro Iwashita et al. “Personalizing Augmented Flashcards Towards Long-Term Vocabulary Learning”. In: Proceedings of the Augmented Humans International Conference 2024. AHs ’24. 2024, Pages 263 –266. DOI: 10.1145/3652920.3653045. URL: https://doi.org/10.1145/3652920.3653045.

[19] Hang Jiang et al. “Leveraging Large Language Models for Learning Complex Legal Concepts through Storytelling”. In: The 62nd Annual Meeting of the Association for Computational Linguistics. 2024. URL: https://arxiv.org/abs/2402.17019.

[20] Nikhita Joshi and Daniel Vogel. “Constrained Highlighting in a Document Reader can Improve Reading Comprehension”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI ’24. 2024, Article No.: 893,Pages 1–10. DOI: 10.1145/3613904.3642314. URL: https://doi.org/10.1145/3613904.3642314.

[21] Gloria Ashiya Katuka et al. “Integrating Natural Language Processing in Middle School Science Classrooms: An Experience Report”. In: Proceedings of the 55th ACM Technical Symposium on Computer Science Education V. 1. SIGCSE 2024. 2024, Pages 639 –645. DOI: 10.1145/3626252.3630881. URL: https://dl.acm.org/doi/10.1145/3626252.3630881.

[22] Joanne Leong et al. “Putting Things into Context: Generative AI-Enabled Context Personalization for Vocabulary Learning Improves Learning Motivation”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI ’24. 2024, Article No. 677, pp. 1–15. DOI: 10.1145/3613904.3642393. URL: https://doi.org/10.1145/3613904.3642393.

[23] Pengfei Liu et al. “Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing”. In: ACM Computing Surveys, Volume 55, Issue 9. 2023, Article No.: 195, Pages 1 –35. DOI:10.1145/3560815.
URL: https://dl.acm.org/doi/10.1145/3560815.

[24] Wenhan Lyu et al. “Evaluating the Effectiveness of LLMs in Introductory Computer Science Education: A Semester-Long Field Study”. In: Proceedings of the Eleventh ACM Conference on Learning @ Scale. L@S ’24. 2024, Pages 63 –74. DOI: 10.1145/3657604.3662036. URL: https://dl.acm.org/doi/10.1145/3657604.3662036.

[25] Reza Hadi Mogavi et al. “When Gamification Spoils Your Learning: A Qualitative Case Study of Gamification Misuse in a Language-Learning App”. In: Proceedings of the Ninth ACM Conference on Learning. L@S ’22. 2022, Pages 175–188. DOI:10.1145/3491140.3528274. URL: https://doi.org/10.1145/3491140.3528274.

[26] Usha Nagasundram et al. “From Connectivity to Dependency: Smartphones in Higher Education Mobile Learning (M-learning)”. In: Proceedings of the 2024 the 16th International Conference on Education Technology and Computers. ICETC’24. 2025, Pages 407 –412. DOI: 10.1145/3702163.3702447. URL: https://doi.org/10.1145/3702163.3702447.

[27] OpenAI. “GPT-4 Technical Report”. URL: https://arxiv.org/abs/2303.08774.

[28] OpenAI. “Hello GPT-4o”. URL: https://openai.com/index/hello-gpt-4o/.

[29] Hyanghee Park and Daehwan Ahn. “The Promise and Peril of ChatGPT in Higher Education: Opportunities, Challenges, and Design Implications”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI ’24. 2024, Article No.: 271, Pages 1 –21. DOI: 10.1145/3613904.3642785. URL: https://doi.org/10.1145/3613904.3642785.

[30] Zhenhui Peng et al. “Storyfier: Exploring Vocabulary Learning Support with Text Generation Models”. In: Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology. UIST ’23. 2023, Article No.: 46, Pages 1 –16. DOI: 10.1145/3586183.3606786. URL: https://doi.org/10.1145/3586183.3606786.

[31] Anh Thi Van Pham and Linh Thi Thuy Le. “Students’ Perceptions towards Duolingo App as a Microlearning Tool”. In: Proceedings of the 2023 5th World Symposium on Software Engineering. WSSE ’23. 2023, Pages 213 –217. DOI: 10.1145/3631991.363202. URL: https://doi.org/10.1145/3631991.363202.

[32] Rolf Reber, Elizabeth A Canning, and Judith M Harackiewicz. “Personalized education to increase interest”. In: Current directions in psychological science 27, 6 (2018). 2018, pp. 449–454. URL: https://s3.wp.wsu.edu/uploads/sites/2445/2019/09/Reber-Canning-Harackiewicz-2018-CDPS.pdf.

[33] Mary Jane C. Samonte et al. “MemoRecall: A Senior Friendly Web Application for Holistic Intervention of Early Memory Lapses in Geriatric Individuals”. In: Proceeding of the 2024 6th International Conference on Information Technology and Computer Communications. ITCC ’24. 2025, Pages 50 –56. DOI:10.1145/3704391.3704399. URL: https://doi.org/10.1145/3704391.3704399.

[34] Mei Tan and Hari Subramonyam. “More than Model Documentation: Uncovering Teachers’ Bespoke Information Needs for Informed Classroom Integration of ChatGPT”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI ’24. 2024, Article No.: 269, Pages 1 –19. DOI: 10.1145/3613904.3642592. URL: https://doi.org/10.1145/3613904.3642592.

[35] Lev Tankelevitch et al. “The Metacognitive Demands and Opportunities of Generative AI”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI’24. 2024, Article No.: 680, Pages 1 –24. DOI: 10.1145 / 3613904 . 3642902. URL: https://doi.org/10.1145/3613904.3642902.

[36] Kathryn T Wissman, Katherine A Rawson, and Mary Pyc. “How and when do students use flashcards?” In: Memory, vol. 20, no. 6, 2012, pp. 568–579. DOI: 10.1080/09658211.2012.687052. URL: https://doi.org/10.1080/09658211.2012.687052.

[37] Rizky Asih Wulandari, Ruli Hafidah, and Adriani Rahma Pudyaningtyas. “The Effect of Augmented Reality (AR) Flashcard on Early Literacy of Early Childhood”. In: Proceedings of the 4th International Conference on Learning Innovation and Quality Education. ICLIQE 2020. 2021, Article No.: 103, Pages 1 – 5. DOI: 10.1145/3452144.3452246. URL: https://doi.org/10.1145/3452144.3452246.

[38] J.D. Zamfirescu-Pereira et al. “Why Johnny Can’t Prompt: How Non-AI Experts Try (and Fail) to Design LLM Prompts”. In: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. CHI ’23. 2023, Article No.:437, Pages 1 –21. DOI: 10.1145/3544548.3581388. URL: https://doi.org/10.1145/3544548.3581388.

[39] Chao Zhang et al. “Mathemyths: Leveraging Large Language Models to Teach Mathematical Language through Child-AI Co-Creative Storytelling”. In: Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. CHI ’24. 2024, Article No.: 274, Pages 1 –23. DOI: 10.1145/3613904.3642647. URL: https://doi.org/10.1145/3613904.3642647.

[40] Wenting Zhang et al. “Navigating the Application Challenges of ChatGPT in Education: Promoting Responsible Use and Minimizing Mental Risks”. In: Proceedings of the 2024 9th International Conference on Distance Education and Learning. ICDEL ’24. 2024, Pages 23 –28. DOI: 10.1145/3675812.3675843. URL: https://doi.org/10.1145/3675812.3675843.

簡易檢索 / 詳目顯示

相關論文