New framework X-AddGraph adds explainability to graph anomaly detection

Researchers have developed X-AddGraph, a novel post-hoc explainability framework for AddGraph, a recurrent graph anomaly detection system. This new method provides auditable reasons for anomaly detection scores without compromising the original detector's performance. X-AddGraph utilizes a Dual Spatial-Temporal Attribution (DSTA) mechanism that aligns with AddGraph's architectural components, offering insights into both spatial relationships and temporal dynamics. AI

IMPACT Enhances trust and auditability in AI-driven anomaly detection systems, crucial for regulated environments.

RANK_REASON The cluster contains an academic paper detailing a new method for explainability in graph anomaly detection. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• AddGraph
• Dual Spatial-Temporal Attribution
• gated recurrent unit
• graph convolutional network
• UCI Message
• X-AddGraph

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →