Classification and identification of medical insurance fraud: a case-based reasoning approach

Appropriate classification of medical insurance fraud events can not only be effective in preventing and combating fraud, but also greatly improve the utilization of medical resources. Due to the uncertainty inherent in medical insurance fraud, identifying and classifying the fraud are non-trivial tasks. In addition, the selection of classification radius by traditional methods is often highly subjective. To this end, a case-based reasoning (CBR) approach in probabilistic hesitant fuzzy environment and its application to classifying the severity of medical insurance fraud events are investigated in this article. At first, the probabilistic hesitant fuzzy element (PHFE) is regarded as a discrete probability distribution, and its distribution function is defined. On this basis, a distribution discrepancy degree is proposed to make up for the shortage of existing measures between PHFEs. Then, a probabilistic hesitant fuzzy decision-making method based on CBR is proposed, which considers both decision data and the expert’s own knowledge and experience. Finally, the proposed method is used to classify the severity of medical insurance fraud events, and the rationality and superiority of the method are verified by comparative analysis. First published online 15 July 2025