LogRAIL: A Retrieval-Augmented LLM Reverification Layer for Log Anomaly Detection
IEEE Access · 2026 · 14 · 65899-65911
DOI: 10.1109/access.2026.3688834 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Wongwang Choi und 6 weiteren Autor:innen. Die Arbeit nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Korea Institute of Planning and Evaluation for Technology in Food, Agriculture and Forestry, LG Electronics, Korea Institute for Advancement of Technology.
Abstract (Originalsprache)
Android logs vary across devices, build versions, and deployment cycles and may contain missing or out-of-order entries, making reliable anomaly detection difficult. To address this problem, we propose LogRAIL, a two-stage framework in which the first stage selects candidate anomaly windows and the second stage performs reverification. Raw logs are normalized into templates, and a transformer-based sequence classifier processes fixed-length windows of log templates to produce anomaly decisions and scores. A retrieval-augmented large language model inference layer (RAIL) then re-evaluates only windows near the decision threshold and applies either a precision-oriented mode to reduce false positives or a recall-oriented mode to reduce false negatives. The model also provides a concise reason for each decision. In both operating modes, LogRAIL improves F1-score over Stage 1 while enabling controlled precision–recall trade-offs aligned with operational objectives. These results show that LogRAIL provides per-window decision explanations and supporting templates, offering template-based rationales for post-detection review and reporting.
Förderung
- Korea Institute of Planning and Evaluation for Technology in Food, Agriculture and Forestry
- LG Electronics
- Korea Institute for Advancement of Technology
- Chung-Ang University
Zitieren
APA
Choi, W., Park, D., Kim, M., Cho, S., Lee, S., Park, J., & Park, H. (2026). LogRAIL: A Retrieval-Augmented LLM Reverification Layer for Log Anomaly Detection. IEEE Access, 14, 65899-65911. https://doi.org/10.1109/access.2026.3688834
Vancouver
Choi W, Park D, Kim M, Cho S, Lee S, Park J, et al. LogRAIL: A Retrieval-Augmented LLM Reverification Layer for Log Anomaly Detection. IEEE Access. 2026;14:65899-65911. doi:10.1109/access.2026.3688834
BibTeX
@article{Choi2026,
author = {Choi, Wongwang and Park, Donghee and Kim, Myeonggwan and Cho, Subin and Lee, Seonghun and Park, Jaehwa and Park, Ho-Hyun},
title = {LogRAIL: A Retrieval-Augmented LLM Reverification Layer for Log Anomaly Detection},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {65899--65911},
doi = {10.1109/access.2026.3688834}
}
EUVIMED