"FraudRAG-2400: A Synthetic Benchmark for Financial Statement Fraud Detection Using Knowledge Graph-Augmented Retrieval"
Abstract
"FraudRAG-2400 is a reproducible synthetic benchmark of 2,400 annotated financial statements for evaluating fraud detection systems. It contains 1,600 legitimate and 800 fraudulent statements which spans for six document types (bank, credit card, payroll, investment, mortgage, tax) and five fraud categories: balance manipulation (34%), amount digit substitution (26%), date forgery (18%), entity name mismatch (12%), and fabricated transactions (10%). All samples are programmatically generated with institution-specific formatting. The benchmark generation script is included for full reproducibility. Associated with IEEE Transaction and Knowledge Data Engineering submission: FraudRAG \u2014 Knowledge Graph-Augmented Retrieval for Financial Statement Fraud Detection. Code and generation scripts are available at following github path: https:\/\/github.com\/akshaysharma1088\/fraudrag"
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.