| 王丽萍,李立,李燕.AI辅助学术写作分布特征的量化分析——以《中华血管外科杂志》(中英文)为例.编辑学报,2026,38(1):90-95 |
| AI辅助学术写作分布特征的量化分析——以《中华血管外科杂志》(中英文)为例 |
| Quantitative analysis of the distribution characteristics of AI-assisted academic writing:a case study of the Chinese Journal of Vascular Surgery(Chinese and English) |
| |
| DOI：10.16811/j.cnki.1001-4314.2026.01.015 |
| 中文关键词: 人工智能生成内容 论文写作 AI检测 AI辅助率 学术诚信 |
| 英文关键词: generative artificial intelligence technology academic writing AI detection AI-assisted rate academic integrity |
| 基金项目:*广东省高水平科技期刊建设优秀人才项目(2025A1212150012,2025A1212150013)；中国高校科技期刊研究会万方数据学术诚信与版权专项基金项目(CUJS2025-CX-ZZ02) |
|
| 摘要点击次数: 644 |
| 全文下载次数: 369 |
| 中文摘要: |
| 人工智能生成内容(AIGC)技术作为一种辅助创新的工具,深刻影响了学术写作模式。人工智能(AI)技术为学术写作带来辅助和便捷的同时,如果使用不当,也可能引发学术诚信问题。检测AI辅助率一定程度上可以鉴别文章的原创性和真实性,识别AI生成的内容,帮助期刊制定更科学合理的稿件筛选标准,确保发表论文的学术价值与质量。本文基于《中华血管外科杂志》(中英文)2023 年 7 月至 2025 年 3 月期间 228 篇初稿的AI辅助率数据,从 AI 辅助率的区间分布、时间变化趋势及文献类型差异3个维度,系统揭示了 AI 技术在学术写作中的应用特征与潜在问题。研究发现,AI 辅助写作已呈常态化趋势,AI的使用引起学术诚信风险升级,不同文献类型对 AI 的容斥度存在差异。对此,本文提出加强科研人员AI使用的规范和边界意识,建立分层监管制度和强制AI使用声明制度,根据文献类型制定差异化的AI使用规范,以推动 AI 技术在学术产出中的合理应用,维护学术诚信与知识生产的健康生态。 |
| 英文摘要: |
| As an auxiliary tool for innovation, artificial intelligence generated content(AIGC) technology has profoundly influenced academic writing.While AI technology brings assistance and convenience to academic writing, improper use may also lead to academic integrity issues. Detecting the AI-assisted rate to some extent can identify the originality and authenticity of articles, recognize AI-generated content, and help journals establish more scientific and reasonable manuscript screening standards to ensure the academic value and quality of published papers. Based on the AI-assisted rate data of 228 preliminary drafts of the Chinese Journal of Vascular Surgery(Chinese and English) from July 2023 to March 2025, this paper systematically reveals the application characteristics and potential problems of AI technology in academic writing from three dimensions:the interval distribution of AI-assisted rates, temporal change trends, and differences in literature types. The study finds that AI-assisted writing has become a regular trend, the academic integrity risks caused by AI use have escalated, and there are differences in the acceptance of AI among different literature types. We propose strengthening the awareness of norms and boundaries for AI use among researchers, establishing a hierarchical supervision system and a mandatory AI use declaration system, and formulating differentiated AI use norms based on literature types. This aims to promote the reasonable application of AI technology in academic output and maintain the healthy ecosystem of academic integrity and knowledge production. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|
