EUVIMEDEUVIMED entdecken
Crossref

LogDLR: Unsupervised Cross-System Log Anomaly Detection Through Domain-Invariant Latent Representation

Junwei Zhou, Shaowen Ying, Shulan Wang, Dongdong Zhao, Jianwen Xiang, Kaitai Liang, Peng Liu

IEEE Transactions on Dependable and Secure Computing · 2025-07 · 22 (4) · 4456-4471

DOI: 10.1109/tdsc.2025.3548050 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Dependable and Secure Computing, 2025-07. Verfasst von Junwei Zhou und 6 weiteren Autor:innen. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 53 Literaturangaben. Gefördert von: National Key Research and Development Program of China, Key Research and Development Program of Hubei Province, NSF.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Log anomaly detection aims to discover abnormal events from massive log data to ensure the security and reliability of software systems. However, due to the heterogeneity of log formats and syntaxes across different systems, existing log anomaly detection methods often need to be designed and trained for specific systems, lacking generalization ability. To address this challenge, we propose LogDLR, a novel unsupervised cross-system log anomaly detection method. The core idea of LogDLR is to use universal sentence embeddings and a Transformer-based autoencoder to extract domain-invariant latent representations from log entries, which can effectively adapt to log format changes and capture semantic information and dependencies in log sequences. To obtain domain-invariant latent representations, we adopt a domain-adversarial training strategy, introducing a domain discriminator that competes with the Transformer-based encoder through a gradient reversal layer, forcing the encoder to learn shared knowledge between different system logs. Finally, the Transformer-based decoder detects anomalies based on the domain-invariant representations obtained by the encoder. We evaluate LogDLR in simulated cross-system scenarios using three publicly available log datasets. The experimental results show that LogDLR can handle heterogeneous logs effectively in cross-system scenarios and achieve efficient and accurate anomaly detection on both source and target systems.

Förderung

  • National Key Research and Development Program of China
  • Key Research and Development Program of Hubei Province
  • NSF

Zitieren

APA

Zhou, J., Ying, S., Wang, S., Zhao, D., Xiang, J., Liang, K., & Liu, P. (2025). LogDLR: Unsupervised Cross-System Log Anomaly Detection Through Domain-Invariant Latent Representation. IEEE Transactions on Dependable and Secure Computing, 22(4), 4456-4471. https://doi.org/10.1109/tdsc.2025.3548050

Vancouver

Zhou J, Ying S, Wang S, Zhao D, Xiang J, Liang K, et al. LogDLR: Unsupervised Cross-System Log Anomaly Detection Through Domain-Invariant Latent Representation. IEEE Transactions on Dependable and Secure Computing. 2025;22(4):4456-4471. doi:10.1109/tdsc.2025.3548050

BibTeX
@article{Zhou2025,
  author = {Zhou, Junwei and Ying, Shaowen and Wang, Shulan and Zhao, Dongdong and Xiang, Jianwen and Liang, Kaitai and Liu, Peng},
  title = {LogDLR: Unsupervised Cross-System Log Anomaly Detection Through Domain-Invariant Latent Representation},
  journal = {IEEE Transactions on Dependable and Secure Computing},
  year = {2025},
  volume = {22},
  number = {4},
  pages = {4456--4471},
  doi = {10.1109/tdsc.2025.3548050}
}
Kostenlos starten →