实验报告抄袭检测系统的研究  

Research on plagiarism detection system of experiment report

作　　者：于海浩[1] 汪伟[1] 黄成哲[1] 孙栩[1] YU Haihao;WANG Wei;HUANG Chengzhe;SUN Xu(College of Computer Science and Technology,Heilongjiang Institute of Technology,Harbin 150050,China)

机构地区：[1]黑龙江工程学院计算机科学与技术学院,哈尔滨150050

出　　处：《黑龙江工程学院学报》2021年第3期1-4,共4页Journal of Heilongjiang Institute of Technology

基　　金：黑龙江省自然科学基金项目(F2018029)。

摘　　要：现有的抄袭检测模型主要是基于启发式的源检索查询生成方法,没有根据抄袭检测任务本身进行建模。文中设计实现了一个基于统计机器学习方法的实验报告抄袭检测系统,该系统使用基于逻辑回归模型从可疑文本片段对中提取词法特征、语法特征、语义特征和结构特征,并通过自动获得的抄袭语料进行不断训练和模型更新。实践证明,通过基于统计机器学习的抄袭检测算法的运用,提高了抄袭检测系统在高模糊抄袭上的性能,为提高实践教育质量提供保障。The existing plagiarism detection software mainly uses a set of detection standards to determine plagiarism,and cannot be dynamically adjusted based on the plagiarism detection task.In this paper,a plagiarism detection system of experiemt report based on statistical machine learning is proposed.This system uses Logical Regression model to combine lexicon features,syntax features,semantics features and structure features extracted from suspicious text fragments pairs.The model is continuously trained and updated through automatically obtained plagiarism corpus.Experiments show the plagiarism detection algorithm based on statistical machine learning is used,the performance of the plagiarism detection system on highly ambiguous plagiarism is improved,which provides a guarantee for improving the quality of practical education.

关 键 词：抄袭 抄袭检测 抄袭语料 实践教学 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]