探討在情境化與個人化下以智能QAC機制促進EFL寫作;Smart QAC Mechanism to Facilitate EFL Writing in Contextualization and Personalization

NCU Institutional Repository > 資訊電機學院 > 網路學習科技研究所 > 博碩士論文 > Item 987654321/92749

請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/92749

題名:	探討在情境化與個人化下以智能QAC機制促進EFL寫作;Smart QAC Mechanism to Facilitate EFL Writing in Contextualization and Personalization
作者:	賴昱夫;Lai, Yu-Fu
貢獻者:	網路學習科技研究所
關鍵詞:	EFL 寫作;真實語境;辨識技術;智慧回饋;生成式人工智能;EFL Writing;Authentic context;recognition technology;smart feedback;generative-AI
日期:	2023-07-12
上傳時間:	2023-10-04 16:09:57 (UTC+8)
出版者:	國立中央大學
摘要:	大多數與EFL寫作相關的研究通常只使用文法錯誤檢查功能來幫助EFL學習者檢查寫作錯誤。然而，這是遠遠不夠的，因為EFL學習者必須學會如何創造更有意義的寫作內容，特別是在真實語境中善用學習者自身周遭的環境。因此，我們開發了一個應用程式，Ubiquitous English (UEnglish)，以其圖像轉文字辨識技術(ITR)提供單字和來自真實語境中的圖片敘述，以及其生成式人工智能產生富有意義的提問(Q)和澄清(C)，來使得EFL學習者產出更多寫作內容。此外，學習者必須在收到來自人工智能的澄清前回答其提問。因此，我們提出了包含三個不同面向的智能提問-回答-釐清(QAC)機制來協助學習者寫作，那些面向包括了原因、交流以及組織。本研究將35名參與者分成兩組，實驗組(EG)有19名學生，控制組(CG)有16名學生，實驗組加入了智能QAC機制輔助，控制組則沒有智能QAC機制輔助。在這項研究中，本次實驗共進行了六周，而我們採用了定量分析。後測結果及學習者行為分析顯示，使用ITR技術和智能QAC機制的實驗組與沒有使用的控制組有明顯差異。此外，實驗組在作業中能寫出更多富有意義的詞句。我們還發現，實驗組學生的原因及組織面向的回答能有效提升他們的後測成績。此外，實驗組學生感受到他們通過了智能QAC機制的輔助獲得了更多有用的建議進而提升了他們的寫作品質。因此，智能QAC機制可以有效促進EFL學習者在真實語境中的寫作能力。;Most studies of EFL writing usually used grammar checking to help EFL learners to check writing errors. However, it is not enough since EFL learners have to learn how to create more meaningful content, particularly using their surroundings in authentic contexts. Therefore, we develop one App, Ubiquitous English (UEnglish), with recognition technology with Image-to-Text Recognition (ITR) texts to provide the vocabulary and description from authentic pictures, and generative-AI that can provide meaningful questions (Q) and clarifications (C) to trigger EFL learners to write more. In addition, EFL learners need to answer the question from AI before they receive the clarification. Hence, we proposed Smart Questioning-Answering-Clarification (QAC) mechanism including three dimensions such as reasoning, communication, and organization to help EFL writing in authentic contexts. A total of 35 participants were assigned into two groups, experimental groups (EG) with 19 learners and control groups (CG) with 16 learners with/without Smart QAC mechanism support, respectively. In this study, the experiment was conducted over six weeks and we used quantitative analysis methods. The results revealed that the EG with ITR-texts and Smart QAC had a significant difference with CG in the learning behaviors and posttest. Furthermore, EG could write more meaningful words in the assignments. In detail, the EG’s answers of the reasoning and organization dimensions were helpful to enhance their writing in the posttest. In addition, EG learners felt that the Smart QAC mechanism helps them to gain more useful suggestions and enhance their writing. Therefore, the Smart QAC mechanism could facilitate EFL learners to enhance their EFL writing in authentic contexts.
顯示於類別:	[網路學習科技研究所 ] 博碩士論文

文件中的檔案:

檔案	描述	大小	格式	瀏覽次數
index.html		0Kb	HTML	47	檢視/開啟

在NCUIR中所有的資料項目都受到原著作權保護.

社群 sharing

資料載入中.....