A framework extracts a latent state machine from logs, induces a multi-table relational schema, and uses it as a generative prior to create synthetic data that augments real logs for better anomaly detection.
Loghub: A large collection of system log datasets towards automated log analytics
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
dataset 1
citation-polarity summary
years
2026 6verdicts
UNVERDICTED 6roles
dataset 1polarities
use dataset 1representative citing papers
LogMILP enables both bag-level anomaly detection and instance-level localization in logs using only bag-level labels via prototype-guided structural modeling and counterfactual perturbation regularization.
IOCRegex-gen automates IOC-to-regex conversion with LLMs via group-aware grouping and multi-stage validation, reporting 99.1% hit rate and 0.8% false-positive rate on 3000+ CTI reports and 2400 ground-truth strings.
AnomalyGen synthesizes realistic labeled log sequences from source code via Log-Oriented Control Flow Graphs and LLM CoT verification to boost F1 scores of 12 anomaly detection models on HDFS and Zookeeper.
A benchmark finds prompt-based LLMs achieve F1 scores of 0.82-0.91 for log anomaly detection in zero-shot settings without any labeled training data, while fine-tuned transformers reach 0.96-0.99.
NLLog rewrites log templates into WHO-WHAT-SEVERITY sentences, applies TF-IDF pooling and tree-ensemble classification with TreeSHAP back-projection, and reports better performance than two reproduced baselines on HDFS and BGL with low false positives on commodity hardware.
citing papers explorer
-
State Machine Guided Multi-Relational Synthetic Data from Logs for Anomaly Detection
A framework extracts a latent state machine from logs, induces a multi-table relational schema, and uses it as a generative prior to create synthetic data that augments real logs for better anomaly detection.