China Knowledge Network has set a threshold of 5% for the sensitivity of this duplicate checking system. In terms of paragraphs, plagiarism or quotations below 5% cannot be detected. So how to detect plagiarism in a paper? The condition of hownet paper detection is that 13 words with continuous similarity or plagiarism will be marked as red, but the prerequisite must be met: that is, the total number of words in a document you quoted or plagiarized and the number of words detected in each paragraph must reach more than 5% before it can be detected as red.
PaperPass's comparative fingerprint database contains more than 90 million academic journals and dissertations, and a database of more than 654.38 billion Internet pages. Fingerprint database refers to the extracted key feature information of an article, not the full text. The papers included in PP detection system are not as comprehensive as HowNet. In order to prevent some reference papers from being included in the database, there is a special self-built database for comparison and reference papers uploading and testing.
When Paperpass first came out, it was said to be stricter than HowNet, because it was repeated for eight consecutive words!
Since it is stricter than HowNet, why not HowNet? Because PP detection system paper database update is slow!
The comparison resources of VIP detection system include: China periodical discussion database, master's degree discussion database, university characteristic discussion database, self-built characteristic discussion database, Internet data resources, etc. VIP claims to be consistent with HowNet's data comprehensiveness and algorithm, or even stricter (with reservations).
This is from the national library education.