Multi-factor Acquisition of Global State based Log-anomaly Detection
Zhiyan Lu, Xiaoyu Zeng · 2023
As an important component of the Fourth Industrial Revolution, Industrial Internet of Things is highly demanded for its stability and confidentiality in production. Discovering anomalies in system operation through log mining is a widely used approach. However, existing methods often remove certain phrases in log messages to gain compressed information, which compromises the integrity of log messages and affects log feature extraction and anomaly detection. To address this issue, we propose a log anomaly detection method called MAD, which utilizes the information of log variable correlation objects to optimize the system's global state analysis reflected in the log. Through experiments, we demonstrate that this method can achieve satisfactory results while consuming fewer resources compared to conventional methods.