学术文献相似性检测比对资源应用分析与建设策略探究——基于万方检测系统的实证分析  

The Exploration of Resource Construction Strategy in Academic Literature Similarity Detection System:Empirical Analysis Based on Wanfang Detection

作　　者：蒋勇青[1] 刘芳[1] 于洋 

机构地区：[1]中国科学技术信息研究所,北京100038

出　　处：《数字图书馆论坛》2017年第12期39-44,共6页Digital Library Forum

摘　　要：比对资源库建设是学术文献相似性检测系统建设中的重要环节。本文通过重点研究实际检测统计数据并分析引用或非典型性引用甚至可能的抄袭文献的特征,为检测系统比对资源建设提供可行策略与发展思路。从2016年向万方数据论文相似性检测系统送检的学术文献中随机抽样650万篇,对抽样送检文献检测结果中共计845 889篇被引用或非典型性引用甚至可能的抄袭文献从文献类型、被相似频率、发表时间等维度进行数据分析。调查发现,比对数据中的学位论文引用或非典型性引用甚至可能的抄袭次数最多,发表或公开时间在2—10年的比对数据引用或非典型性引用甚至可能的抄袭次数最多。此次研究为将来学术文献相似性检测系统的持续建设与优化提供全新的思路与方法借鉴。The data base construction is an important part of the academic literature similarity detection system development. This paper gathered the information of 845 889 similarity articles, acquired from 6.5 million srandom sampling journal papers, dissertations and other type literatures submitted in the Wanfangdata’s Similarity Detection system in 2016. The paper summarized and evaluated the statistical data in the aspects of literature types, plagiarism numbers and published time. Results showed that, the dissertations had the highest cited or plagiarized number among all the literature types, at the same time, the papers published in 2 to10 years had the most cited or plagiarized times.The research provided new ideas and methods for the further development of the academic literature similarity detection system.

关 键 词：文献检测 比对资源建设 学术不端 

分 类 号：G353.1[文化科学—情报学]