TMLC adopts adaptive multilevel fingerprint (AMLFP) feature detection technology independently developed by CNKI, which has the characteristics of fast detection speed, high accuracy, high recall rate and strong anti-interference. Support the detection of chapters, paragraphs and sentences at all levels; Support document deformation detection such as document rewriting and multi-document combination; Support the detection of academic misconduct in long documents such as graduate dissertations and book monographs.

For any document that needs to be detected, the system first carries out hierarchical processing, and creates fingerprints according to the levels of chapters, paragraphs and sentences. At the same time, it also uses the same technology to create fingerprint indexes for the comparative documents in the resource library.

Extended data:

Relevant requirements for paper duplicate checking:

1. If the paper submitted by the user is in MS Word format and the document directory is generated according to MS Word format, the detection system will automatically identify the chapters of the paper and display the contents of the paper according to the actual chapter information of the paper.

2. If there is no obvious chapter information in the dissertation, or the dissertation is not in MS Word format, the system will automatically segment the dissertation according to each paragraph of 1000 words or more, and display it according to the segmentation results.

3. The number of overlapping words refers to the number of words in the overlapping part of the paper after comparing this chapter with the comparative literature. Regardless of the reproduction ratio, the more overlapping words, the greater the possibility of academic misconduct.

Baidu Encyclopedia-Academic Misconduct Detection System