The repetition rate has nothing to do with the number of words, but the similarity is related to the number and length of words cited by the author. The larger the length of the author's citation, the higher the repetition rate will be detected, and the more words in the article, the less the part cited by the author, so it will not be detected that the repetition rate is too high. On the contrary, articles with few words but large citations are bound to be found out. The "duplicate checking rate" of extended data is also called repetition rate, similarity rate and plagiarism rate. In fact, every word has the same meaning, but the name is different. The so-called duplicate checking rate refers to the quantification of the similarity between the results detected by the paper detection system and others' documents. The plagiarism rates of commonly used paper detection systems are: Wanfang Paper Detection System, HowNet Paper Detection System, Paperpass Paper Pass, Weipu Paper, Paperrater and so on. According to the different requirements of each school, different systems will be selected for testing. For example, XX school requires testing the system of HowNet, so students do not need to repeatedly test other systems. According to the system standard of HowNet, because the algorithm and database comparison of each detection system are different, the detection result in HowNet is 10%, which may be 30% in paperpass.