Financial Crime World

Fraud Detection Tools and Techniques in Belgium: A Boost for VLAIO’s Subsidy Program

Introduction

Flanders Innovation & Entrepreneurship (VLAIO) has deployed a cutting-edge fraud detection system to curb fraud and streamline its subsidy program. Developed by Devoteam, the tool provides a risk score for each submitted request for subsidies under three key aid measures: SME e-wallet (KMOP), Ecology Premium Plus (EP-Plus), and COVID-19-related support.

The System

The system is designed to help VLAIO inspectors target high-risk subsidiaries more efficiently, saving time and resources. By analyzing data from multiple sources, including VLAIO’s own database and the Central Database for Enterprises (KBO), the tool identifies potential fraudsters and provides a comprehensive overview of each applicant’s risk profile.

Benefits

  • Saves time and resources by targeting high-risk applicants
  • Provides a comprehensive overview of each applicant’s risk profile
  • Enables inspectors to identify potential fraudsters quickly and focus on targeted inspections

Methods Used

The fraud detection system uses a combination of underlying methods, including:

  • Community detection
  • Rule-based methods
  • Supervised machine learning
  • Anomaly detection

These methods are integrated into a custom-built Power BI dashboard that inspectors can use to visualize the data and gain insights into each applicant’s risk profile.

Data Security

One of the key challenges faced by VLAIO was ensuring the security of sensitive data used in the system. The tool is designed to handle data on sole proprietorships, which includes information about individual citizens, requiring strict confidentiality measures throughout the process.

Azure Environment

The system has been built within the Azure environment and its code is written in Python, utilizing other Azure components such as:

  • Azure Data Factory (ADF) for automated processing
  • Power BI dashboard for visualizing data

Results

Since deployment, inspectors have reported significant time savings and improved readability of data. The custom-built dashboard has also served as a centralized information hub, eliminating the need to access multiple systems to gather relevant information.

Key Takeaways

  • Time savings for inspectors
  • Improved readability of data
  • Centralized information hub

Conclusion

VLAIO’s fraud detection system represents a significant innovation in the field of subsidy management in Belgium, demonstrating the power of technology in detecting and preventing fraudulent activities while streamlining administrative processes.