An Examination of Insurance Fraud in Korea: An AHP Approach

As insurance fraud becomes one of the most serious white-collar crimes in the capitalist society, insurance industry has spent billions of dollars to deter and detect the insurance fraud. This is the issue in Korea, too. Since so called 'IMF economic crisis' of Korea, insurance fraud cases have increased dramatically. Several approaches have been adopted to analyze the fraud problem by researchers, this paper employs analytic hierarchy process (AHP) and extracts expert's knowledge. The three objectives of this study are (1) to explore automobile insurance fraud in Korea, (2) establish the hierarchical structure of insurance fraud identification, and (3) to compare the relative importance of variables among different models. This study analyzed the automobile bodily injury claim data from December of 1996 to November of 1998. Three level hierarchy of insurance fraud indication was developed. Out of 15 indicators, Accident Cause turned out to be the most important indicator (22.9)%), following Accident Type (11%), Number of Past Accidents (10.3%), Multiple Policies (9.2%) and so on. Fraud detection power is believed to be improved by integrating AHP aided expert's model into data mining models.