Winnowing算法和动态规划算法在作业剽窃检测中的应用和比较  

Application and Comparison of the Winnowing Algorithm and the Dynamic Programming Algorithm in Assignment Plagiarism Detection

作　　者：张量[1] 刘秀敏[1] 刘秀娟[2] 

机构地区：[1]杭州师范大学信息科学与工程学院,浙江杭州310018 [2]嘉兴学院,浙江嘉兴314001

出　　处：《计算机工程与科学》2009年第6期147-149,共3页Computer Engineering & Science

摘　　要：信息数字化后使抄袭、剽窃变得更加容易了,为了杜绝学生作业中出现的抄袭现象,急需高效的剽窃检测技术,帮助教师对学生作业的抄袭情况实施监督检测。本文分别用具有除噪和过滤功能的Winnowing算法和动态规划算法对学生作业进行剽窃检测。通过对文档间相似度的对比,实现了对作业剽窃程度的检测。实验结果表明,Winnowing算法更加有效、更加可靠。Plagiarism becomes more easily for the informational digitalization, even in the students＇ assignments. In order to completely eradicate the phenomenon of plagiarism,there is an urgency of effective techniques of plagiarism detection to help teacher monitor plagiarism in students＇ assignments. This paper applies the techniques of the Winnowing algorithm,which has the functions of filtering and eradicating disturbance, and the dynamic programming algorithm to detect the students＇ assignments. We use the similarity of two texts to detect the degree of plagiarism between assignments. The results show that the Winnowing algorithm is more effective and reliable.

关 键 词：剽窃检测 Winnowing算法 动态规划算法 最长公共子序列 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]