【摘要】 文档复制检测技术是保护知识产权和提高信息检索效率的有效有段。目前,关于英文的复制检测技术已经趋向成熟,但关于中文的复制检测技术才刚刚起步。在此提出基于文档指纹的中文复制检测方法,改进了正向最大匹配分词方法并建立了相应的语料库机制,提出k-words分解句子的方法;定义了两篇文档同粒度指纹矩阵之间的匹配运算。实验结果表明,所提出的方法能高效地进行中文文档复制检测。更多还原

【Abstract】 The technique of copy detection is an efficient approach to protect intellectual property and improve efficiency of information retrieval.Currently,many research efforts have been devoted to the development of this technique for English.However,copy detection technique for Chinese just strides the first step.A new method of word segmentation and a dictionary mechanism correspond to it are designed.An efficient method for syncopating sentence,k-words,is presented.The concepts of describing fingerprints with matrix and matching fingerprint from different granularity are proposed.The match operation between fingerprints’ matrixes which have the same granularity is defined.Finally,experiments show that the method proposed is efficient in copy detection of Chinese.更多还原