Framework for the identification of fraudulent health insurance claims using association rule mining

Abstract

Deliberate cheating by concealing and omitting facts while claiming from health insurance providers is considered as one of fraudulent activities in the health insurance domain which has led to significant amount of monetary loss to the providers. In view of the above, careful scanning of the submitted claim documents need to be conducted by the insurance companies in order to spot any discrepancy that indicates fraud. For this purpose, manual detection is neither easy nor practical as the claim documents received are plentiful and for diverse medical treatments. Hence, this paper shares the initial stage of our study which is aimed to propose an approach for detecting fraudulent health insurance claims by identifying correlation or association between some of the attributes on the claim documents. With the application of a data mining technique of association rules, this study advocates that the successful determination of correlated attributes can adequately address the discrepancies of data in fraudulent claims and thus reduce fraud in health insurance.