A Plagiarism Detection Method Based on Learning Behavior Analysis

Wen-jun TANG, Du ZOU, Ling ZHANG


Plagiarist detection is a complicate topic in plagiarism detection area. Most existing algorithms can only compare similarity between assignments, but cannot detect plagiarist. This paper designed a homework assessment model based on the South China University of Technology e-learning plagiarism detection module. By analyzing students' learning behavior data collected from the teaching platform, a ranking of the possibilities of plagiarism in students' work is obtained, which provides the basis for judging the plagiarist. And comparing the determination of plagiarism with the actual investigation results, the accuracy of plagiarist determination using the model is significantly improved. The model has been used on the e-learning platform, which provided an effective way for teachers to evaluate assignments.


Learning behavior analysis, Plagiarism detection, Similarity, Clustering, E-learning


Full Text: