基于特征向量的中文文献相似度评价

Evaluating Similarities of Chinese Literatures Based on Characteristic Vectors

科技文献抄袭现象伴随着科学技术的发展时有发生,这严重损害了文献原作者,也对科技文献的严肃性提出了挑战。本文利用分词技术提取文献特征向量,并结合动态规划算法对文献的相似度给出具体评价,针对不同抄袭的现象,发现其中存在的规律,具体问题具体分析。最后给出实际实验结果,为文献评审提供参考。

With the development of science and technology,the phenomenon of plagiarizing scientific literatures occurs frequently,which not only severely injures the original author of the literature,but also challenges the seriousness of the scientific literatures.In this paper,the word segmentation technologies are used to extract the literature characteristic vectors,and combined with the dynamic programming algorithm,the similarities of the literatures are evaluated specifically.For different plagiarizing phenomena,we find there are different laws.Specific problems need to make specific analysis.Finally,the actual experimental results are given to be used for reference in literature appraisal.