抄袭剽窃论文识别研究综述

Literature Review on Copy and Plagiarism Detections

抄袭剽窃论文的识别是知识产权保护中一项重要的内容,已有众多的识别方法和系统.本文从抄袭剽窃的定义、文本的表示（向量空间模型、广义向量空间模型、隐性语义索引模型）、文本相似度的研究内容、文本相似度的计算方法（基于统计学的计算方法和基于语义理解的计算方法）、数字指纹和词频统计两大类技术和方法和抄袭剽窃识别系统等方面为基本思路,对该领域中已提出的主要研究方案进行了分类阐述和比较分析,总结了其最新研究进展,为下一步的研究提出了新的课题和设想.

Copy and plagiarism detection is emerging as one of the primary research areas in intellectual property protection. Many plagiarism detection methods and systems have been proposed. The paper summaries this research field from some points of view, such as the definition of copy and plagiarism, text representation（ such as, Vector Space Model, Generalized Vector Space Model, Latent Semantic Index）, research content of text similarity, computation method of text similarity（ such as one computation method based on statistics, another computation method based on semantic comprehension）, the two main techniques and methods, namely, finger printing and word frequency, and detection systems. At the end of the paper, some difficulties have to overcome in the future are pointed out, and directions to study are given.