Anomaly detection on software log based on Temporal Memory. (October 2021)

• Record Type: Journal Article
• Authors: Hirakawa, Rin; Uchida, Hironori; Nakano, Asato; Tominaga, Keitaro; Nakatoh, Yoshihisa
• Is Part Of: Computers & electrical engineering. Volume 95(2021)
• Publication Date: 2021-10
• Subjects: Anomaly detection -- Software log -- Temporal memory -- Unsupervised learning -- Time series pattern -- Defect analysis; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.107433
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: For complex system failures, it is necessary to investigate the cause using runtime log data as a clue. The amount of log data output by systems is becoming increasingly large, and there is a need for an automatic log analysis method that can extract only the log data related to the failure. In this study, we propose TM-LAD, a method for automatically detecting log patterns that are different from normal system operation. In our experiments, we compared the performance of anomaly detection with seven existing methods using the loglizer benchmark. The results show that the proposed method has the smallest F-measure of 0.01 standard deviation in the benchmark, and is more robust to changes in the amount of training data than the other methods.