Analytics for Insurance Fraud Detection: An Empirical Study

Abstract

Automobile insurance fraud is a global problem. Handling fraud manually has always been costly for insurance companies. Data analytics can play a crucial role in fraud detection and can aid insurance companies to identify fraud. Typically, there are easily more than thirty variables that are used for the fraud analysis. This paper proposes to determine which variables are significant for fraud detection and to provide a framework for the insurance fraud detection.

Further, this paper illustrates the business value of data analytics for insurance fraud detection using an empirical study and demonstrates that through a few business rules, the insurance company can accurately identify fraudulent claims which can most likely reduce costs and increase profitability for the company.