生成式大语言模型赋能研究生《学术写作能力》课程的改革路径——以北京交通大学课程改革为例
Reform Paths for the Postgraduate Course “Academic Writing Ability” Empowered by Generative Large Language Models—A Case Study of Curriculum Reform at Beijing Jiaotong University
DOI: 10.12677/ae.2026.16112508, PDF,   
作者: 师 海*:北京交通大学土木建筑工程学院,北京;轨道工程北京市重点实验室,北京;冀向瑾:北京交通大学土木建筑工程学院,北京
关键词: 生成式大语言模型;学术写作课程;教学变革;变革路径;AI赋能;Generative Large Language Models; Academic Writing Course; Teaching Reform; Reform Path; AI Empowerment
摘要: 生成式大语言模型(Large Language Models, LLMs)的发展为《学术写作能力》课程的教学带来革命性契机。文章以北京交通大学研究生《学术英语写作》课程实践为基础,结合实践中学术英语写作中心建设、AI (Artificial Intelligence)写作应用案例等,系统剖析LLMs赋能课程的多维机遇,分析应用中的结构性挑战,并提出“技术–教学–伦理”三位一体的系统化应对框架。通过北京交通大学的“AI + 写作”实践表明:LLMs可通过AI赋能、模式创新和能力升级三重路径重构教学方式,但需破解学术诚信风险、AI发展与现有能力落差、专业与AI适配不足等核心矛盾。因此,构建适配研究生教学的模式、实施分层教学改革、健全伦理规范体系,是实现上述变革的关键路径。
Abstract: The development of generative large language models (LLMs) has brought revolutionary opportunities to the teaching of the “Academic Writing Ability” course. Based on the practice of the postgraduate “Academic English Writing” course at Beijing Jiaotong University, and incorporating initiatives such as the establishment of an academic English writing center and case studies on the application of AI writing, this paper systematically examines the multi-dimensional opportunities of LLMs in empowering the course, analyzes structural challenges in their application, and proposes a systematic tripartite framework addressing “technology - teaching - ethics”. The “AI + Writing” practice at Beijing Jiaotong University demonstrates that LLMs can reshape teaching methods through three pathways: AI empowerment, model innovation, and competency enhancement. However, core challenges such as academic integrity risks, the gap between AI development and existing competencies, and insufficient discipline-specific adaptation of AI must be resolved. Therefore, constructing a postgraduate-teaching-oriented model, implementing tiered teaching reforms, and improving an ethical regulatory system are key pathways to achieving this transformation.
文章引用:师海, 冀向瑾. 生成式大语言模型赋能研究生《学术写作能力》课程的改革路径——以北京交通大学课程改革为例[J]. 教育进展, 2026, 16(11): 1339-1344. https://doi.org/10.12677/ae.2026.16112508

参考文献

[1] 吴鹏飞, 郭鹏, 等. 基于大模型和多源学习资源数据的课程知识图谱建设研究[J]. 中国信息技术教育, 2025(21): 111-114.
[2] 侯永雄, 钟伟森, 等. 人工智能赋能高校创新创业教育高质量发展耦合逻辑与实现路径[J]. 高教学刊, 2024, 10(26): 75-78.
[3] OpenAI Research Team (2022) The Impact of Large Language Models on Academic Writing Efficiency and Quality. Journal of Educational Technology & Society, 25, 124-141.
[4] 董雅清. AI辅助学术英语写作的认知与行为演化——基于质性主题分析的实证研究[J]. 西部学刊, 2026(18): 26-29.
[5] 王海啸. 生成式人工智能在大学英语教学改革中的应用探究——以“通用学术英语写作”课程教学改革实践为例[J]. 外语教育研究前沿, 2024, 7(4): 41-50, 95.
[6] 宋迎春. 大语言模型与论文写作: 问题与对策[J]. 外国语言文学与文化论丛, 2025(0): 179-187.
[7] 谢炜, 黄建业, 郑艳蓉, 等. 基于大语言模型与知识图谱协同推理的辅助专业写作智能体框架研究[J]. 图书情报工作, 1-13.
https://link.cnki.net/urlid/11.1541.G2.20260810.0701.002, 2026-09-26.
[8] 韩雪雯, 车尚锟, 杨梦晴, 等. 多模态数据驱动的Al智能体模式设计[J]. 图书情报工作, 2024, 68(24): 27-37.
[9] 孙丝雨, 侯跃芳, 丁敬达, 等. 基于多源数据的疾病知识图谱构建研究[J]. 情报工程, 2024, 10(4): 3-13.
[10] Le, T., Le, N. and Le, B. (2023) Knowledge Graph Embedding by Relational Rotation and Complex Convolution for Link Prediction. Expert Systems with Applications, 214, Article ID: 119122.
https://doi.org/10.1016/j.eswa.2022.119122
[11] Barratt, D., et al. (2023) The Effect of Long-Term LLM Dependence on Academic Critical Thinking. Higher Education Research & Development, 42, 1689-1708.
[12] 高歌, 丁慧玲. 理工科高校学术英语写作的学科适配性问题与对策[J]. 北京交通大学学报(社会科学版), 2024, 23(1): 152-160.
[13] 李雪平, 李棱雪. 基于BOPPPS的翻转课堂教学模式应用与研究——以土木工程材料课程为例[J]. 科教导刊, 2024(33): 116-118.