.png){kind=icon}
.png){kind=icon}
Return
Manod: A multi-modal anomaly detection framework for distributed system
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neunet.2025.107999.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Pioneered a advanced approach by integrating log and metric data for multimodal anomaly detection. • Applied a hierarchical graph-based encoding method to model inter-metric and intra-metric correlations in time-series. • Applied prompt-based pre-trained language models for log modeling. • Implemented a novel cross-modal fusion technique to effectively align and unite diverse data modalities. • Demonstrated the practical applicability of multimodal anomaly detection in complex distributed systems.
Keywords:
Distributed system
Anomaly detection
Multimodal learning
Time series analysis
Log modeling
Deep learning
Journal
IF:
6.3
Papers:
7.8K
Citations:
3.0W