识别科技论文学术不端2法  

Two methods for identifying academic misconduct in scientific papers

作　　者：武晖 孟超 师琅 徐杰[2] WU Hui;MENG Chao;SHI Lang;XU Jie(Editorial Department of Journal of Xi’an Polytechnic University,710048,Xi’an,China;不详)

机构地区：[1]西安工程大学学报编辑部,西安710048 [2]中山大学学报(医学科学版)编辑部,广州510080

出　　处：《编辑学报》2020年第3期295-298,302,共5页Acta Editologica

基　　金：西安工程大学高等教育研究项目(11299000314)。

摘　　要：目前还没有合适的自动检测方法来识别图表的学术不端行为,主要靠人工判断。通过收集、梳理图片学术不端的类型,借助实例论述拉伸变形和局部修剪、移位和截取、旋转和裁剪、重复使用等图表学术不端的鉴别方法。运用OSID工具包中的语音录制和上传研究佐证材料等识别科技论文的学术不端问题。实践结果表明,这2种检测方法能够防范科技论文中存在的学术不端行为。There is no proper automatic detection method to identify the academic misconduct of charts except for manual detection. By collecting and reviewing the types of academic misconduct of image,we discussed the identification methods of academic misconduct,which involve drawing deformation and local pruning,shifting and intercepting,rotating and cutting,reusing.Moreover,the voice recording and uploading research supporting materials in the OSID toolkit were used to identify the academic misconduct of scientific papers. Our results show that these two detection methods can prevent the academic misconduct in scientific papers.

关 键 词：科技论文 学术不端 图片篡改 学术期刊 科研诚信 OSID工具包 

分 类 号：G232[文化科学]