Agentic data science with the Data Agent Kit
cloud.google.com·6h ago
AI Research
arXiv CS.AI·Aug 26·5 min read
LEMMA-RCA: A Large Multi-modal Multi-domain Dataset for Root Cause Analysis
TL;DR
LEMMA-RCA is a comprehensive dataset designed for root cause analysis (RCA) across various domains, including IT and OT systems. It enables fair evaluation of RCA techniques, promoting advancements in system reliability and performance.
✦ Why It Matters
Start using LEMMA-RCA today to enhance your RCA models and improve system reliability.
Key Takeaways
- Utilize LEMMA-RCA for training RCA models in diverse domains.
- Evaluate your RCA techniques against established baseline methods.
- Explore fault scenarios in IT and OT systems for practical insights.
- Contribute to the open-source community by sharing findings from LEMMA-RCA.
Full Summary
Root cause analysis (RCA) is essential for improving the reliability of complex systems, yet the field has been limited by a lack of large, open-source datasets. LEMMA-RCA addresses this gap by providing a substantial dataset that encompasses real-world fault scenarios from diverse domains such as Information Technology (IT) and Operational Technology (OT).
The dataset includes hundreds of system entities and supports various modalities, making it versatile for different RCA tasks. Six baseline methods were evaluated on LEMMA-RCA, demonstrating its effectiveness in both offline and online model settings.
Results indicate that LEMMA-RCA facilitates a fair evaluation of RCA techniques, encouraging the development of more robust solutions. The dataset is publicly accessible, promoting collaboration and innovation in the field.