【目的】 探究编辑部对作者投稿进行第一次学术不端检测(以下简称“初检”)后,稿件中学术不端检测结果发生动态变化的成因,探索初检后学术不端的主动控制措施。【方法】 从笔者的经历以及相关学者的研究文献入手,分析造成稿件在初检后复制比发生动态变化的原因。【结果】 稿件在初检后学术不端检测结果发生动态变化的原因是稿件之间的抄袭,并且抄袭与被抄袭的论文都未进入数据库,其中一篇稿件刊发并且进入数据库后,其他稿件的学术不端即刻表现出来;分布的主要群体是师生、学术团队、同事和学术研究合作者;大多数学术不端内容在投稿前就已经存在,少数为作者投稿后修改期间植入;检测软件自身算法缺陷以及比对数据库更新周期滞后都是造成学术不端的检测结果动态变化的原因。【结论】 为了大幅度降低初检后学术不端的出现概率,编辑部应当实行优先发表、缩短稿件的出版周期、刊发前复检和采用两种主流学术不端检测软件交叉检测来稿等措施;数据库公司应当保证数据的正确性,缩短数据库的更新周期;软件开发公司应当优化软件算法,最大程度地健全、完备数据库。
[Purposes] This paper aims to explore the causes of the dynamic changes after the first academic misconduct detection (preliminary detection) by editorial offices, and put forward measures to control new academic misconduct after the preliminary detection. [Methods] Based on the writer's own experience and literature research, the causes of dynamic changes of manuscripts plagiarism after the preliminary detection were analyzed. [Findings] The dynamic changes after the preliminary detection are due to plagiarism. When either of the original or plagiaristic manuscripts, which have not been put into database before the preliminary detection, are published and put into database as literatures, the others will be detected as academic misconducts. This situation mainly happens among teachers and students, academic teams, colleagues, and academic research collaborators. Most academic misconducts exist before submission, and some authors' modifications after submission also lead to academic misconducts. Other causes include arithmetic bugs of software and update cycle lag of caparison database. [Conclusions] The editorial offices are advised to publish in advance, shorten publication cycle, detect the manuscript once more before publication, and use two major academic misconduct detectors to check manuscripts. The database companies should ensure the correctness of data and shorten the database update cycle. The software development companies need to optimize software algorithm and perfect the database. These measures can significantly reduce academic misconducts after preliminary detection.
Academic misconduct literature check system,