中国科技期刊研究 ›› 2021, Vol. 32 ›› Issue (5): 563-570. doi: 10.11946/cjstp.202101060019

• 学术不端防范专题 • 上一篇    下一篇

防范图片学术不端的举措研究

孙力炜(), 贺郝钰, 迟秀丽, 侯春梅   

  1. 中国科学院西北生态环境资源研究院文献情报中心, 甘肃省兰州市天水中路8号 730000
  • 收稿日期:2021-01-06 修回日期:2021-03-22 出版日期:2021-05-15 发布日期:2021-05-15
  • 作者简介:孙力炜(ORCID:0000-0001-7826-9972),硕士,编辑,E-mail: sunlw@llas.ac.cn;|贺郝钰,硕士,副编审,编辑部主任;|迟秀丽,学士,副编审;|侯春梅,硕士,编审,期刊编辑出版部主任。
  • 基金资助:
    中国科学院自然科学期刊编辑研究会2020年资助课题“新技术在科技期刊图片学术不端防范中的应用”(YJH-WT-008)

Strategies for preventing image academic misconduct in scienpngic journals

SUN Liwei(), HE Haoyu, CHI Xiuli, HOU Chunmei   

  1. Lanzhou Information Center, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, 8 Middle Tianshui Road, Lanzhou 730000, China
  • Received:2021-01-06 Revised:2021-03-22 Online:2021-05-15 Published:2021-05-15

摘要:

【目的】 通过对国内外防范图片学术不端中采用的方法、技术以及软件等的特征及其实用性进行调查分析,以期为国内科技期刊在选取合适的图片学术不端防范举措时提供参考。【方法】 采用文献调研法和对比分析法,比较国内外图片学术不端的防范举措及其实用性。【结果】 采用“制定论文原始数据提交标准及其处理指南规范作者行为”+“编辑审核”+“专家评审”等方式,可有效防范基于数据驱动软件生成图片的学术不端行为。对于设备拍摄图片的学术不端的防范尚存在较多难点,现已有查询图像属性和源代码、JPEGsnoop、MagicEXIF、iPlagiarism等方法和软件能够为编辑在鉴别图片篡改行为中提供一定的帮助;已有平台和算法实现了图片剽窃检测,但仅有少量出版机构能够加以应用。【结论】 国内科技期刊界亟须制定广泛认可的数据提交体系和标准,并鼓励有实力的出版机构建立能够存储并管理科技期刊论文数据的仓储,同时积极联合算法研究者、软件开发者以及大型数据库研发能够高效自动化检测图片篡改和剽窃行为的系统或平台,帮助期刊进一步提升防范学术不端行为的能力和效率。

关键词: 学术不端, 图片剽窃, 图片伪造, 图片篡改

Abstract:

[Purposes] This paper investigated and analyzed the characteristics and practicability of the methods, technologies, and software for preventing image academic misconduct in China and abroad, which is expected to serve as a reference for scienpngic journals in China in selecting suitable countermeasures. [Methods] Through literature research and network analysis, we compared the strategies for preventing image academic misconduct in China and abroad. [Findings] For the image fabrication via software, the effective countermeasures are formulating original data standard and processing guidelines, checking by editors, and expert review. Despite the difficulty in idenpngying and preventing image falsification, there are some methods and software, such as checking image properties and source code, JPEGsnoop, MagicEXIF, and iPlagiarism. Platforms and algorithms for image plagiarism detection have been developed, which, however, are rarely used by publishers. [Conclusions] It is urgent for scienpngic journals in China to establish a commonly accepted data submission system and standard and the construction of a data warehouse for storing and managing paper data of scienpngic journals is preferred. It suggested that they should actively cooperate with algorithm and software developers and large databases to develop systems or platforms for efficiently and automatically detecting image falsification and plagiarism. Thereby, the ability and efficiency to prevent academic misconduct will be further boosted.

Key words: Academic misconduct, Image plagiarism, Image fabrication, Image falsification