An anomaly detection of learning behaviour data based on discrete Markov chain. (20th December 2022)

• Record Type: Journal Article
• Authors: Li, Dahui; Qu, Peng; Jin, Tao; Chen, Changchun; Bai, Yunfei
• Is Part Of: International journal of continuing engineering education and lifelong learning. Volume 33:Number 1(2023)
• Page Start: 69
• Page End: 83
• Publication Date: 2022-12-20
• Subjects: discrete Markov chain -- learning behaviour data -- support vector machine -- redundancy; Engineering -- Study and teaching -- Periodicals; 620.00715
• Languages: English
• ISSNs: 1560-4624

Abstract

In order to overcome the problems of large anomaly detection error and long detection time in traditional learning behaviour data anomaly detection methods, this paper proposes a learning behaviour data anomaly detection method based on discrete Markov chain. This method analyses the types of learning behaviour data, and determines the influencing factors of learning behaviour data. With the help of support vector machine, the data extraction range is determined, and the data redundancy is determined to complete the data pre-processing. This paper analyses the basic principle of discrete Markov chain, constructs the discrete Markov chain model, and completes the detection of abnormal learning behaviour data. The experimental results show that the maximum detection error of the proposed method is about 2%, and the detection time is always less than 2.5 s.