IFGDS: An Interactive Fraud Groups Detection System for Medicare Claims Data

Jing Yu, Zhenyang Yu, Kaiming Zhan, Fan Wu, Bo Yin, Lei Duan

Abstract: Safeguarding medicare fund security is a key concern in the medical and healthcare fields. Effectively detecting the fraud groups from massive amounts of medicare claims data is technically challenging. Therefore, we developed an interactive fraud groups detection system named IFGDS. It can screen out suspicious claims data from a large volume of medicare claims data and then detect fraud groups with “ganging up” medical visit behaviors and abnormal treatment characteristics. IFGDS provides abundant visual monitoring and analysis tools as well as a user-friendly interaction experience. It facilitates users to view patients-related information. Moreover, IFGDS has high cross-platform-portability to handle a comprehensive range of data resources. In practice, IFGDS has been applied to medicare regulatory scenarios, and has successfully recovered medicare fund among several cities.