系統識別號 | U0002-2008202313290400 |
---|---|
DOI | 10.6846/tku202300584 |
論文名稱(中文) | 基於 ChatGPT 與 Azure 雲端平台建構一個口說及文法學習系統~以日文為例 |
論文名稱(英文) | Constructing a Spoken and Grammar Learning System Based on ChatGPT and the Azure Cloud Platform: A Case Study of Japanese |
第三語言論文名稱 | |
校院名稱 | 淡江大學 |
系所名稱(中文) | 資訊工程學系碩士班 |
系所名稱(英文) | Department of Computer Science and Information Engineering |
外國學位學校名稱 | |
外國學位學院名稱 | |
外國學位研究所名稱 | |
學年度 | 111 |
學期 | 2 |
出版年 | 112 |
研究生(中文) | 王品皓 |
研究生(英文) | Pin-Hao Wang |
學號 | 610410556 |
學位類別 | 碩士 |
語言別 | 繁體中文 |
第二語言別 | |
口試日期 | 2023-06-09 |
論文頁數 | 35頁 |
口試委員 |
指導教授
-
張志勇(cychang@mail.tku.edu.tw)
共同指導教授 - 郭經華(chkuo@mail.tku.edu.tw) 口試委員 - 廖文華(whliao@ntub.edu.tw) 口試委員 - 游國忠(133742@mail.tku.edu.tw) 口試委員 - 張志勇(cychang@mail.tku.edu.tw) |
關鍵字(中) |
ChatGPT Azure平台 語音轉文字 文字轉語音 日語學習 LineBot 情境學習 |
關鍵字(英) |
Speech-to-text Text-to-speech Contextual Learning Sentence Grammar Checking Sentence Fluency Accuracy Analysis |
第三語言關鍵字 | |
學科別分類 | |
中文摘要 |
傳統的語言學習環境中,學習者常常需要依賴教師或其他擅長該語言的人提供即時糾正。然而,這種即時糾正的機會並不總是容易獲得,尤其當學習者在自主學習時。在自主學習的情況下,即使學習者有意無意地發音、用字或文法錯誤,由於沒有人及時糾正,這些錯誤可能會變成日後的發音及構句,影響日語交流的語言能力。因此,了解自己的發音、用字及文法是否準確,才能從而提高日語口語表達能力,並且能夠更全面地評估自身的語言能力。此外,對於日本語學習者而言,情境學習也是一個重要的學習方式。傳統的語言學習方式常常著重於學習單詞和文法規則,但這樣的學習方式可能無法真正培養學習者在日常對話中流利使用語言的能力。情境學習可以讓學習者在實際情境中進行日語溝通練習,學習如何運用語言表達自己的意思。本研究「基於ChatGPT 與Azure雲端平台建構一個口說及文法學習系統~以日文為例」,提出結合ChatGPT和Azure雲端平台的支援功能,構建一個口說及文法學習系統,並提供兩個平台:LineBot和Web端。LineBot平台主要是用來實現個人練習功能,學習者可以透過LineBot進行日文口說練習。並透過ChatGPT的自然語言生成能力,直接以口語形式向LineBot提問,進行日文句子的發音、用字及文法練習。系統會在練習過程中即時糾正錯誤,以提升日語口語表達能力。Web端則提供情境學習功能。學習者進入Web端時,可以選擇不同情境的對話,Web端將根據情境設計好的劇情,使用ChatGPT進行對話模擬,讓學習者在實際情境中進行日語對話練習。同時,Web端也將利用ChatGPT和Azure雲端平台進行語法難易度分析和問答檢查技術,以及即時評估學習者的文法正確性,並根據日文語法標準給予評分和回饋,幫助學習者了解自己在文法應用上的準確度,並改進錯誤。 |
英文摘要 |
In traditional language learning environments, learners often rely on teachers or proficient speakers of the language to provide real-time corrections. However, these opportunities for immediate corrections may not always be readily available, especially in self-directed learning scenarios. When learners make pronunciation errors, whether intentionally or unintentionally, without timely corrections, these mistakes may become ingrained habits affecting the fluency of Japanese communication. Therefore, understanding the accuracy of one's pronunciation is crucial for improving Japanese oral expression and obtaining a comprehensive assessment of language proficiency.Additionally, contextual learning is essential for Japanese language learners. Traditional language learning methods often focus on vocabulary and grammar rules, but they may not fully develop learners' ability to use the language fluently in everyday conversations. Contextual learning allows learners to practice Japanese communication in real-life situations, teaching them how to apply language effectively. This research, "Constructing a Speaking and Grammar Learning System Based on ChatGPT and Azure Cloud Platform - A Case Study in Japanese," proposes the integration of ChatGPT and Azure Cloud Platform to build a speaking and grammar learning system, offering two platforms: LineBot and Web.The LineBot platform provides personalized practice, enabling learners to engage in Japanese speaking exercises. Utilizing ChatGPT's natural language generation capabilities, learners can directly communicate with LineBot in spoken language and practice Japanese sentence pronunciation. The system provides real-time corrections during practice sessions, enhancing Japanese oral expression skills.On the other hand, the Web platform offers contextual learning. Learners can select different dialogue scenarios, such as shopping, dining, or traveling, and engage in Japanese conversation simulations within these contexts. The Web platform utilizes ChatGPT and Azure Cloud Platform for grammar difficulty analysis and question-answer checks, providing instant assessments of learners' grammar accuracy based on Japanese grammar standards and offering feedback for improvement. |
第三語言摘要 | |
論文目次 |
目錄 誌謝 I 目錄 V 圖目錄 VII 表目錄 VIII 第一章 簡介 1 1.1 背景與動機 1 1.2 目的與貢獻 2 第二章 相關研究 9 2.1 單功能系統 9 2.2 多功能系統 10 第三章 背景知識 13 3.1 Azure TextToSpeech 13 3.2 Azure SpeechToText 14 3.3 Azure Speech Pronunciation 15 3.4 OpenAI ChatGPT 16 3.5 google WebkitSpeechRecognition 17 3.6 Flask 框架 18 第四章 系統架構 20 4.1 環境與問題描述 20 4.1.1 欲解決問題 20 4.1.2 目標 20 4.2 整體架構 20 4.2.1 個人練習 22 4.2.2 融合情境導向口說引導與文法學習功能間的整合 25 第五章 實驗分析 28 5.1 實驗設計 28 5.2 效能評估 30 5.3 學習體驗評估 31 第六章 結論 33 參考文獻 34 圖目錄 圖 1 、 無法及時得知發音是否正確 2 圖 2 、透過LineBot即時回應給使用者 3 圖 3 、 系統設計之劇情 4 圖4 、挑選劇情並指定角色 5 圖 5、情境發音練習 6 圖 6、情境會話練習 7 圖7 、 使用Azure TextToSpeech API 14 圖 8 、 使用ChatGPT API 17 圖9 、Flask串接 19 圖10、系統架構圖 21 圖11、個人練習功能 22 圖12、個人練習(練習回答)流程圖 24 圖13、個人練習(練習發音)流程圖 25 圖14、情境學習功能 25 圖15、系統產生的多種劇情影片 26 圖16、情境對話及發音練習流程圖 27 圖17、多語言情境學習回饋表 29 圖18、不同時間下使用與未使用本系統的效能比較折線圖 30 圖19、學習者滿意度圓形圖 32 圖20、系統易用性圓形圖 32 表目錄 表 1 、相關功能研究比較表 12 |
參考文獻 |
參考文獻 [1] Y. Kurniawan, F. A. N. Septiani, E. Halim and G. Bhutkar, "An Idea of Japanese Language Learning for Indonesians via Mobile Application Development," 2022 8th International HCI and UX Conference in Indonesia (CHIuXiD), Bali, Indonesia, 2022, pp. 13-18 [2] J. He, "An Exploratory study on the application of artificial intelligence technology in the teaching of Japanese language in university," 2021 2nd International Conference on Information Science and Education (ICISE-IE), Chongqing, China, 2021, pp. 1454-1457 [3] H. Takagi, S. Xi, M. Kuwahara, K. Ueki and M. Yabuki, "A study on practical communication application for beginners of Japanese," 2021 IEEE 10th Global Conference on Consumer Electronics (GCCE), Kyoto, Japan, 2021, pp. 215-217 [4] H. C. Kurniawan, B. Sitohang and S. A. Rukmono, "Gamification of Mobile-based Japanese Language Shadowing," 2019 International Conference of Artificial Intelligence and Information Technology (ICAIIT), Yogyakarta, Indonesia, 2019, pp. 215-219 [5] A. Gangopadhyay, I. Bardhan, A. Das, N. S. Soman and S. Das, "Democratizing Language Learning using Machine Learning," 2022 56th Annual Conference on Information Sciences and Systems (CISS), Princeton, NJ, USA, 2022, pp. 137-141 [6] N. Shi, Q. Zeng and R. Lee, "Language Chatbot–The Design and Implementation of English Language Transfer Learning Agent Apps," 2020 IEEE 3rd International Conference on Automation, Electronics and Electrical Engineering (AUTEEE), Shenyang, China, 2020, pp. 403-407 [7] V. Suresh, R. Agasthiya, J. Ajay, A. A. Gold and D. Chandru, "AI based Automated Essay Grading System using NLP," 2023 7th International Conference on Intelligent Computing and Control Systems (ICICCS), Madurai, India, 2023, pp. 547-552 [8] Kondo, Mariko, et al. "Phonetic fluency of Japanese learners of English: automatic vs native and non-native assessment." Proceedings of the International Conference on Speech Prosody. Vol. 2020. 2020. [9] Hong, Yeonjung, and Hosung Nam. "Evaluating score reliability of automatic English pronunciation assessment system for education." Stud. Foreign Lang. Educ 35 (2021): 91-104. [10] Moorhouse, Benjamin L., and Andrew M. Beaumont. "Utilizing video conferencing software to teach young language learners in Hong Kong during the COVID‐19 class suspensions." TESOL Journal 11.3 (2020). [11] Mai, Long, and Julie Carson-Berndsen. "Enhancing conversational quality in language learning chatbots: An evaluation of GPT4 for ASR error correction." arXiv preprint arXiv:2307.09744 (2023). [12] Tejedor-García, Cristian, Valentín Cardeñoso-Payo, and David Escudero-Mancebo. "Automatic speech recognition (ASR) systems applied to pronunciation assessment of L2 Spanish for Japanese speakers." Applied Sciences 11.15 (2021): 6695. [13] J. Nundlall and V. Ramnarain-Seetohul, "A Feedback Generation System to Enhance Learning at Primary School," 2020 3rd International Conference on Emerging Trends in Electrical, Electronic and Communications Engineering (ELECOM), Balaclava, Mauritius, 2020, pp. 67-72 [14] Y. Ono, S. Hirokawa, M. Ishihara and M. Yamashiro, "Real time text-based feedback systems from frequency-based feedback to mindmap feedback in foreign language teaching," 2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC), San Diego, CA, USA, 2014, pp. 2150-2153 [15] Liu, Yanqing, et al. "Delightfultts: The microsoft speech synthesis system for blizzard challenge 2021." arXiv preprint arXiv:2110.12612 (2021). [16] Le, Chenyang, et al. "ComSL: A Composite Speech-Language Model for End-to-End Speech-to-Text Translation." arXiv preprint arXiv:2305.14838 (2023). [17] Y. Gong, Z. Chen, I. -H. Chu, P. Chang and J. Glass, "Transformer-Based Multi-Aspect Multi-Granularity Non-Native English Speaker Pronunciation Assessment," ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Singapore, Singapore, 2022, pp. 7262-7266 [18] Grinberg, Miguel. Flask web development: developing web applications with python." O'Reilly Media, Inc.", 2018. [19] Vaswani, Ashish, et al. "Attention is all you need." Advances in neural information processing systems 30 (2017). [20] Su, Bin, et al. "Improving pronunciation assessment via ordinal regression with anchored reference samples." ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2021. |
論文全文使用權限 |
如有問題,歡迎洽詢!
圖書館數位資訊組 (02)2621-5656 轉 2487 或 來信