Filters








3 Hits in 0.83 sec

LogBERT: Log Anomaly Detection via BERT [article]

Haixuan Guo, Shuhan Yuan, Xintao Wu
2021 arXiv   pre-print
In this paper, we propose LogBERT, a self-supervised framework for log anomaly detection based on Bidirectional Encoder Representations from Transformers (BERT).  ...  The experimental results on three log datasets show that LogBERT outperforms state-of-the-art approaches for anomaly detection.  ...  Experimental results on three log datasets have shown that Log-BERT outperforms the state-of-the-art approaches for log anomaly detection.  ... 
arXiv:2103.04475v1 fatcat:egchxl7ghnexbmtd4332eobuje

LogBERT: Log Anomaly Detection via BERT

Haixuan Guo
2021
Overall, LogBERT outperforms the state-of-art models for log anomaly detection.  ...  The proposed model, LogBERT, a BERT-based neural network, can capture the contextual information in log sequences.  ...  In this thesis, we develop a novel deep learning model based on BERT, LogBERT, to detect abnormal log sequences.  ... 
doi:10.26076/0567-4901 fatcat:g7fgorylqjbvzf3weufbkkpoc4

UniLog: Deploy One Model and Specialize it for All Log Analysis Tasks [article]

Yichen Zhu and Weibin Meng and Ying Liu and Shenglin Zhang and Tao Han and Shimin Tao and Dan Pei
2021 arXiv   pre-print
UniLog: Deploy One Model and Specialize it for All Log Analysis Tasks  ...  Logbert: Log anomaly and then evaluate on the Hadoop dataset.  ...  For anomaly log anomaly detection. 5) LogAnomaly [3], the state-of-the- detection, failure prediction, summarization and compression art method on log anomaly detection, combines  ... 
arXiv:2112.03159v1 fatcat:zqiycpha7vd3xmpey4loxbrd3a