The principle of paper duplication checking is that 13 characters are continuous and similar, and the repeated content is included in the repetition rate of the paper. The paper duplicate checking system will process the content layer by layer, and create fingerprints by chapter, paragraph and sentence. When comparing contrast documents in a repository, the same technique is used to create a fingerprint index. After the user's paper is uploaded to the duplicate checking system, the system will automatically check the duplicate of the paper, and the duplicate checking report can be provided to the user after the duplicate checking is completed. The main principle is big data, and the similarity of article content is relatively believed. The main purpose of preventing paper duplication is to improve the efficiency of use, so the principle of paper duplication checking is to make big data before speaking. The duplicate checking system has a huge comparison database, and the paper will find out whether there is duplication and how much it accounts for. If the proportion exceeds the requirements of the school, it needs to be reduced.