随机生成内容引发撤稿的特征解析与治理对策  

Characteristics analysis and governance strategies for retractions caused by randomly generated content

作　　者：王晨阳 褚建勋[1,2] 管铮懿[3,4] 杨晓婷 WANG Chenyang;CHU Jianxun;GUAN Zhengyi;YANG Xiaoting(School of Humanities and Social Sciences,University of Science and Technology of China,1129 Huizhou Avenue,Baohe District,Hefei 230051,China;Institute of Computational Social Science and Media Studies,University of Science and Technology of China,1129 Huizhou Avenue,Baohe District,Hefei 230051,China;National Science Library,Chinese Academy of Sciences,33 Beisihuan Xilu,Zhongguancun,Haidian District,Beijing 100190,China;Department of Information Resources Management,School of Economics and Management,University of Chinese Academy of Sciences,33 Beisihuan Xilu,Zhongguancun,Haidian District,Beijing 100190,China;School of Culture and Tourism,Henan University,85 Minglun Street,Shunhehuizu District,Kaifeng 475000,China)

机构地区：[1]中国科学技术大学人文与社会科学学院,安徽省合肥市230051 [2]中国科学技术大学计算社会科学与融媒体研究所,安徽省合肥市230051 [3]中国科学院文献情报中心,北京市100190 [4]中国科学院大学经济与管理学院信息资源管理系,北京市100190 [5]河南大学文化旅游学院,河南省开封市475000

出　　处：《中国科技期刊研究》2024年第11期1635-1644,共10页Chinese Journal of Scientific and Technical Periodicals

摘　　要：【目的】分析随机生成内容(RandomlyGeneratedContent,RGC)技术引发的撤稿现象特征,为治理RGC撤稿提供参考。【方法】采集撤稿观察数据库的3929篇相关文献,用科学计量学方法分析撤稿时间和时滞分布、国家分布、学科分布、撤稿原因、撤稿主体、文献类型及发表载体等特征。【结果】RGC致撤稿量剧增、时滞缩短;撤稿集中于科学后发国家,如中国、印度;计算机科学等领域是撤稿重灾区;高频撤稿原因属学术不端;学术期刊或出版商、第三方为撤稿主力;撤稿文献以研究论文为主,Hindawi撤稿量最多,撤稿量前10期刊多为混合或金色开放获取巨型期刊。【结论】应明确RGC在学术出版中的适用边界与规范;对撤稿高发学科建立更严格的审查机制;推动建立RGC治理联盟。[Purposes]This study aims to thoroughly analyze the characteristics of retraction phenomena caused by randomly generated content(RGC)technology and provide references for managing retractions related to RGC.[Methods]Data was collected from the Retraction Watch Database,containing 3929 related documents.Bibliometric methods were employed to analyze the basic characteristics of retractions,including retraction timing and delay distribution,country distribution,subject distribution,retraction reasons,retraction initiators,document types,and publication platforms.[Findings]In recent years,the number of retractions due to RGC has surged,and the retraction delay has shortened.Retractions are primarily concentrated in emerging scientific countries,particularly China and India.The fields of computer science,technology,and data science are the"hotspots"for retractions.The most frequently cited reasons for retractions are related to academic misconduct.Academic journals or publishers and third parties are the main initiators of retractions.Retraction documents are mostly research papers,with Hindawi having the highest number of retractions.The top ten journals with the most retractions are predominantly mixed open access(OA)or gold OA journals,many of which are"mega-journals".[Conclusions]Clear boundaries and regulations should be established for the application of RGC technology in academic publishing.Stricter review mechanisms are needed for disciplines with high retraction rates,such as computer science.Journals should strengthen their cooperation with third parties.

关 键 词：随机生成内容 撤稿 科研诚信 生成式人工智能 学术不端 

分 类 号：G23[文化科学]