Hyundai Marine & Fire Insurance Co., Ltd., is ready to deploy a new fraud detection system that has been built with software from SAS, a global provider of business analytics software and services.

Hyundai M&F established a strong foundation for preventing costly fraud by implementing a fraud detection system developed by SAS that not only prevents claims fraud but will help enhance customer satisfaction and improve premium payment processing, thereby protecting its most profitable customers.

Hyundai M&F's SAS fraud detection system combines business rules based on the experience and knowledge of its investigators with model rules generated from data extracted from various IT systems using SAS advanced statistical techniques. The models are applied to insurance claims, delivering results to claims investigators in real time. The system also monitors the performance of these rules, enabling the company to quickly modify existing rules or generate new ones. The predictive models enhance the process of fraud detection so Hyundai M&F can detect insurance fraud not just after it occurs, but prevent a fraudulent claim from being paid. 

"With SAS providing the foundation for advanced insurance fraud management, we can maintain consistent and transparent criteria for investigation, automate the scoring system, and respond swiftly and flexibly to investigations and operations of the organization," explained Cheol-woo Lee, the Hyundai M&F Claim Investigation Department team leader who oversaw the project. "In particular, this will improve the efficiency of investigation personnel who collect and analyze evidence data for suspected insurance fraud, preventing payment of unjustified claims, while protecting honest policyholders from insurance fraud."

The project has been carried out in three areas: a fraud pre-detection system to help investigators judge the possibility of fraudulent activities using statistics-based model rules and business rules; a fraud post-detection system to detect afterward the characteristics of certain groups not detected with the prejudgment model; and a risk mart to generate data for prejudgment and post-judgment models.

Register or login for access to this item and much more

All Digital Insurance content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access