Sign in to view source links and access this dataset
Description
FWA Synthetic Graph v1 is a synthetic knowledge graph modeling federal grant and loan program entities with planted fraud typologies. It contains 925 planted fraud clusters and models operations including applicants, applications, payments, vendors, and agencies. The dataset was created by kvirtue and last updated on Hugging Face in May 2026.
Use Cases
Train Graph Neural Networks for anomaly detection based on planted fraud clusters.
Benchmark identity resolution algorithms using the modeled relationships between entities like applicants and vendors.
Simulate and analyze fraud typologies in federal assistance programs based on the described operational model.
Demonstrate Salesforce Data Cloud capabilities using the synthetic graph structure.
Strengths
Contains 925 carefully planted fraud clusters for model training.
Models a realistic federal assistance ecosystem including applicants, applications, payments, vendors, and agencies.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
Row count and dataset size are unknown, which may limit suitability assessment.
Data is synthetic and may not capture all nuances of real-world fraud patterns.
Provenance
Source
kvirtue on Hugging Face.
Collection Method
Synthetically generated for GNN training and demonstration.
Freshness
Last updated 2026-05-11 01:50:37; freshness should be verified.
License is unknown; terms of use must be verified before application.