A Research Of The Intelligent Audit System Of Medical Insurance Based-on Rule

Recent years, China’s social medical insurance business is flourishing. However, with the enlargement of the participants of medical insurance, the component is becoming more complex. The medical insurance has wasted a lot. The main reason is some of the participants of medical insurance often chase small profit and the medical institution lack the sense of social responsibility.They fraud or misappropriate medical insurance fund illegally, which wastes a lot of medical insurance funding. Therefore, it has been an urgent requirement to use modern information technology to discriminate the illegal behavior efficiently.This thesis conducted a research about the key technologies of the intelligent audit system of social medical insurance, and proposed a solution which is mainly based on rule engine and data mining technology. Hadoop, MVC design pattern, JavaEE framework, XML and WebService technologies are also used in this system.. This system can audit the medical insurance payment automatically in the perspective of medical sense, and allows flexible configurations for medical rules.Rule engine is a technology which matches the fact data according to a series conditions of every rule and decides whether to execute rule’s follow-up actions.The mining of big data is the process to obtain hidden information from mass of data. While the medical insurance units have accumulate a lot of data, the data mining of medical data will contribute to the effective supervision of the medical insurance funding. In this thesis, the two technologies are studied and analyzed. Secondly, this thesis clears the goal of the whole system and puts forward the architecture and technology roadmap of the system. Thirdly, it designs the system in detail and proposes a method to express and store medical rules, realizing the storage and dynamic edition of rules and the automatic generation of rule files. Meanwhile, it realizes the core component: rule engine based on Drools. Fourthly, it uses the parallel FP-Growth algorithm based on MapReduce to gain medical professional knowledge and dig out frequent patterns of medicines for some diseases. Furthernore, it assists medical auditors to formulate medical rules. There are three innovations for this thesis:1.By extracting medical rules from medical auditors’ mind or related files, it proposed a unique structured medical rules expression method according to their characteristics, making it easy to express, store and maintaining the rules.2..With the combination of rule engine technology and the medical insurance audit actual situation, the system can audit the medical insurance payment automatically in the perspective of medical sense. And it achieves the flexible configuration of the rules, while reducing the cost of system maintenance.3.The system can dig out frequent patterns for medicines of some diseases by using of data mining technology and Hadoop, which assists medical auditors to formulate medical rules and helps the hospitals to prescribe medicines. It provides new ways to think of medical insurance audit work.