生成式人工智能在高校教学、学习、科研和行政管理中的应用：指导原则和政策
[美]安云乔 [美]俞智贤 [美]萨达拉·詹姆斯
（北得克萨斯大学学习技术系，登顿 TX76207，美国）
肖俊洪 译
【摘 要】本研究调查了美国排名前50的大学针对生成式人工智能（GenAI）在学术和行政管理中的应用的指导原则。我们采用混合式研究方法，包括话题建模、情感分析和质性主题分析，全面分析这些高校应对GenAI的措施。话题建模归纳了四个核心话题：GenAI在学习和考核中的应用、GenAI在视觉和多模态媒体中的应用、GenAI的安全和伦理考量以及GenAI与学术诚信。情感分析发现，这些不同类型的高校对GenAI持非常积极的态度，但是面向教师和面向学生的指导原则存在显著差异。质性主题分析印证了上述分析结果并进一步显示94％的高校教师的指导原则强调必须因课程而异制定相应政策并清楚告知学生。主题分析还发现学术诚信和隐私担忧是反复出现的主题，这点与话题建模的“安全和伦理考量”一致。研究结果表明高等教育领域应迅速出台GenAI的应用指导原则，同时强调必须根据不同对象灵活制定相应政策以应对这种技术所带来的机会和挑战。
【关键词】学术诚信；人工智能；ChatGPT；生成式人工智能；高等教育；机构指导原则；政策
Investigating the Higher Education Institutions’ Guidelines and Policies Regarding the Use of Generative AI in Teaching, Learning, Research, and Administration
Yunjo An, Ji Hyun Yu and Shadarra James
(Department of Learning Technologies, University of North Texas, Denton TX76207, USA)
Abstract: This study examined the guidelines issued by the top 50 U.S. universities regarding the use of Generative AI (GenAI) in academic and administrative activities. Employing a mixed methods approach, the research combined topic modeling, sentiment analysis, and qualitative thematic analysis to provide a comprehensive understanding of institutional responses to GenAI. Topic modeling identified four core topics: Integration of GenAI in Learning and Assessment, GenAI in Visual and Multimodal Media, Security and Ethical Considerations in GenAI, and GenAI in Academic Integrity. These themes were further explored through sentiment analysis, which revealed highly positive attitudes towards GenAI across all institution types, with significant differences between faculty and student-targeted guidelines. Qualitative thematic analysis corroborated these findings and provided deeper insights, revealing that 94% of universities had faculty guidelines emphasizing the importance of establishing and communicating course-specific GenAI policies. This analysis also highlighted recurring themes such as academic integrity and privacy concerns, which aligned with the security and ethical considerations identified in the topic modeling. The study highlights the rapid evolution of GenAI guidelines in higher education and the need for flexible, stakeholder-specific policies that address both the opportunities and challenges presented by this technology.
Keywords: academic integrity; artificial intelligence; ChatGPT; generative AI; higher education; institutional guidelines; policies
下载： 
生成式人工智能在高校教学、学习、科研和行政管理中的应用：指导原则核政策.pdf
