计算领域的论文数据共享与可重复性问题分析  

Analysis of Data Sharing and Reproducibility Problem in Computing Field

作　　者：于倩倩[1,2,3] 孟银涛 钱力[1,2,3] 刘会洲[5] Yu Qianqian;Meng Yintao;Qian Li;Liu Huizhou(National Science Library,Chinese Academy of Sciences,Beijing 100190;Department of Information Resources Management,School of Economics and Management,University of Chinese Academy of Sciences,Beijing 100190;Key Laboratory of New Publishing and Knowledge Services for Scholarly Journals,Beijing 100190;University of International Relations Library,Beijing 100091;Institute of Process Engineering,Chinese Academy of Sciences,Beijing 100190)

机构地区：[1]中国科学院文献情报中心,北京100190 [2]中国科学院大学经济与管理学院信息资源管理系,北京100190 [3]国家新闻出版署学术期刊新型出版与知识服务重点实验室,北京100090 [4]国际关系学院图书馆,北京100091 [5]中国科学院过程工程研究所,北京100190

出　　处：《图书情报工作》2024年第17期3-15,共13页Library and Information Service

基　　金：中国科学院文献情报中心青年骨干人才项目“面向科技文献内容挖掘的通用实体语料库建设研究”(项目编号:E3550201)研究成果之一。

摘　　要：[目的/意义]计算可重复是可靠和可信研究的基石,对计算领域的论文数据共享政策、数据可用性和可重复性方法进行调查分析,为促进数据共享及解决计算可重复问题提供参考。[方法/过程]采用网站调研法和内容分析法,分析期刊、会议的数据共享政策;采用网络爬虫法,获取期刊论文数据的可用性声明,对数据可用性进行分析;并梳理和总结计算可重复的方法。[结果/结论]计算领域大多数期刊、超半数会议有数据共享政策,但数据共享态度的强度还有待提升。越是高水平的期刊或会议,越可能有数据共享政策。相较于期刊关注数据共享问题,会议更关注计算可重复问题。数据可用性声明促进了数据共享,但数据共享实践与数据共享政策要求还存在差距。计算可重复方法包括鼓励或要求数据共享、专家审查、设置奖励、设置论文提交清单、征集可重复性论文等。[Purpose/Significance]Computational reproducibility is the cornerstone of reliable and credible research.To investigate and analyze data policies,data availability and reproducibility methods of journals and conferences in computing field can provide references for promoting data sharing and solving computational reproducibility issues.[Method/Process]With web survey research and content analysis method,it analyzed journal data policies and conference data policies.It used web crawler to obtain data availability statement of journal article and analyze the current situation of data availability.Then,it summarized computational reproducibility methods.[Result/Conclusion]Most journals and more than half of the conferences in computing field have data policies,but the intensity of data sharing attitudes still needs to be improved.Journals or conferences with higher level are more likely to have data policies.Compared with journals focusing on data sharing issues,conferences pay more attention to the problem of computational reproducibility.The data availability statement promotes data sharing,but there is still a gap between data sharing practices and data sharing policy requirements.Computational reproducibility methods include data sharing,expert review,setting rewards,paper submission checklist,calling for reproducibility papers and so on.

关 键 词：期刊论文 会议论文 数据共享 计算可重复 

分 类 号：G250[文化科学—图书馆学]