First, if some papers are in English, you can add notes at the back, each sentence can be rewritten to varying degrees, or you can directly translate English into Chinese, similar to HowNet, PaperRight and other systems that can check duplicates in Chinese and English. Although many foreign language papers can be detected, it is impossible to detect many languages in a paper. Therefore, we can start with foreign languages.
Second, if the number of Chinese papers has far exceeded the number of words required by the school, the method of keyword deletion and replacement can be adopted. You can also use semantic extension to comprehensively add the omitted subject and predicate in the original sentence. Even one more word may help you reduce the repetition rate of 1%, so don't underestimate these details.
The third kind: the special content in the paper, such as the data part. You can convert words into tables, and tables are basically impossible to copy. Words become graphics, tables become graphics, which are clear at a glance and will never be detected as repeated plagiarism.
Fourth: None of the above is applicable. Use this method when it has no effect. Primitive law, what is primitive law? It won't really make you scratch your head to think about how to form an opinion and how to discuss it. This original method is based on reading the contents of the cited documents and being familiar with them, so that you can express your unique views. In other words, you can express other people's papers in your own words, which can minimize the plagiarism rate of papers.