Benchmark RadarTSAI-MetaFraud
TSAI-MetaFraud is a multimodal, multi-task benchmark dataset for fraud detection in virtual economies, integrating behavioral, transactional, and graph-structured data. It defines tasks including fraud detection, node classification, temporal link prediction, and weakly supervised fraud detection, with baseline evaluations.
- Released
- 2026-07-10
- Readiness
- Paper only
- Primary field
- Finance & Economics
Why it matters
Fraud in metaverse ecosystems is a new challenge that combines behavioral and financial data. TSAI-MetaFraud provides a unified benchmark to advance multimodal learning and fraud analytics in virtual economies, filling a gap in existing datasets.
Motivation
The emergence of metaverse platforms has created virtual economies that introduce new challenges related to fraud, bot activity, and illicit financial behavior.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.