Medicare Fraud Gang Discovery Based on Community Discovery Algorithms

With the increasing seriousness of medical insurance fraud, the demand for anti-fraud has gradually increased. In order to seek higher fraud amounts, fraud is often committed in the form of gangs. Therefore, the main research in this paper is to use the community discovery algorithm to mine the fraud of health insurance gangs. There are bound to be profit-seeking scenarios for gang fraud. For various fraud scenarios, we abstract it into the form of a graph, take the insured as a node, and build a variety of relationships between the insured, and then use the graph structure and various relationships to conduct community found, and finally fused all the results. The overall work of this paper can be summarized into three parts. One is to construct a graph of the medical treatment process and medical insurance utilization of a designated retail hospital in a city; The similarity of the graph is reconstructed, and then the community discovery algorithm is used; third, in the final gang mining result, the method of taking the intersection of multi-graph results is adopted and the result is obtained. Experiments have shown that in the field of medical insurance gang fraud, these ideas have indeed achieved good results in the modularity index, and the final mining results are indeed suspicious to a certain extent under the evaluation of the medical teacher, which provides the follow-up work.