Research On Block-chain Technology And Social Network Analysis Based Healthcare Insurance Fraud Identification Method

With the gradual improvement of China’s healthcare insurance system,more and more residents are enjoying the benefits of healthcare insurance and no longer suffer from the problem of "expensive medical treatment".However,the phenomenon of healthcare insurance fraud has become increasingly serious in recent years,with reports of typical cases emerging,showing a trend of hidden methods,large amounts of money,complex cases and groups of perpetrators.Health insurance fraud has not only increased the burden on the insurance industry,but has also become a serious threat to the social security system.In this context,the use of new technologies for fraud identification will help maintain the viability of the health insurance industry and support the operation of the social security system.This paper proposes a framework for information exchange based on block-chain distributed storage technology and social network analysis.After receiving a claim submitted by an insured person,the insurer uses social network analytic to perform an initial fraud identification of the claim.Depending on the details of the claim,the help of relevant nodes is then sought within this framework for validation.Once the nodes have reached consensus under the practical Byzantine consensus mechanism,the claim records are included in the historical database and the local ledger of each node is updated.The empirical part first cleans the data and performs descriptive statistical analysis of the data to initially look for characteristics of suspicious claims submitted by suspected fraudsters.A fraud score is calculated for each claim using the Bi-Rank algorithm.A binary logistic regression model is constructed using the fraud score with the amount class variable and tested to verify feasibility.Possible problems in health insurance fraud identification from a data processing perspective are presented.Finally,advanced experiences in health insurance fraud prevention in developed countries are summarized and recommendations are made at four levels.The novelties of this paper include a blockchain-based health insurance claim validation framework and the application of social network analysis to health insurance fraud identification.Social network analysis performs well on this dataset as the data used in the study shows a concentration of visits by suspected fraudsters.