大語言模型(LLM)驅動筆譯教學:基於認知負荷理論的文化因素提示詞設計研究  

LLM-Driven Written Translation Teaching:A Study on the Design of Culturally Informed Prompts Based on Cognitive Load Theory

作　　者：翟秋蘭 ZHAI Qiulan(Guangzhou College of Applied Science and Technology,China)

机构地区：[1]廣州應用科技學院

出　　处：《人文与社科亚太学刊》2025年第1期104-113,共10页Asia-Pacific Journal of Humanities and Social Sciences

基　　金：2024年广东省教育科学规划课题(高等教育专项)「三全育人」視域下廣東英語專業學生講好灣區故事路徑研究(專案編號:2024GXJK418)。

摘　　要：大語言模型的運用為外語教育和翻譯教學帶來了全方位的變革、前所未有的機遇和挑戰。本文立足於ChatGPT和通義等國內外等大語言模型驅動筆譯課堂,基於認知負荷理論,提出了文化因素提示詞設計的三大原則:第一,避免直线型提问;第二,採取倒金字塔式層級提問方式,深挖文化要素,刻意增加學生的有益負荷;第三,教師應充分發揮主觀能動性,對LLM持批判態度,及時反思生成內容,促進學生的正向認知。同時,通過對ChatGPT和通義等國內外等大語言模型對相同提示詞生成的回答,指出大語言模型驅動下的筆譯教學中,教師應如何做出教學反思;研究旨在為大語言模型驅動筆譯教學做出實踐參考。The integration of large language models has profoundly transformed foreign language education and translation teaching,presenting unprecedented opportunitics and challenges.Grounded in Cognitive Load Theory,this study examines translation classrooms powered by large language models such as ChatGPT and Tongyi and thereby proposes three key principles for designing culturally informed prompts to address translation challenges arising from culural differences.Through a comparative analysis of responses generated by these models to identical prompts,the study elucidates their strengths and limitations.These insights provide actionable recommendations for educators to retlect critically on their instructional strategies.This rescarch contributes practical guidance for integrating large language models into translation teaching,aiming to enhance pedagogical effectiveness and lcarner outcomes.

关 键 词：大語言模型(LLM) 筆譯教學 文化因素提示詞 

分 类 号：H319[语言文字—英语]