Optimization And Application Of Rule Execution Based On Medical Insurance Business

In recent years,with the promotion of medical insurance at the national level,it is increasingly important to prevent medical insurance fraud and protect the stability and safety of medical insurance funds.At present,the main direction of medical insurance is to use information technology to build a reasonable regulatory system,and to use big data technology to identify abnormal data and behavior from massive medical insurance data.This thesis enriches the existing rule base by analyzing and mining the historical medical insurance data,and constructs a comprehensive audit system for the medical insurance business by using the rule engine as the core method.At the same time,in order to meet the needs of massive data audit,the idea of MapReduce model is adopted,and a parallel execution scheme based on existing rule engine is proposed,which effectively improves the audit ability of the system.Through the above methods,a set of solutions for mining new rules,applying rules,and optimizing rule execution performance is formed.The main work includes the following aspects:Firstly,in the aspect of medical insurance rules mining,this thesis analyzes the problems of medical insurance rules mining,and preprocess and feature selection for the medical insurance expense data set.Focus on the problem of data imbalance of the dataset,and compare the performance of multiple data balance strategies in the medical insurance dataset.Finally,the Ripper algorithm is used to induct medical insurance rules,and compared with other rule learning algorithms.Secondly,in the aspect of optimization of rule execution,this thesis based on the existing rule engine Drools,combined with the idea of MapReduce model,a rule parallel execution plan for medical insurance auditing business is proposed.Specifically,the design ideas of the scheme are elaborated in the way of sub-rules splitting and distribution,the distribution of fact data,and the reduction of intermediate results.In this process,the characteristics of the Rete algorithm are further optimized.Finally,the performance of the prototype system and the stand-alone system based on the scheme is compared to prove the feasibility of the scheme.Finally,in the design and implementation of the medical insurance audit system,this thesis analyzes the system requirements,and propose the overall design and implementation architecture of the system.For the detailed functional module design,the design and implementation of core modules such as rule management,rule flow management,audit service and rule monitoring are introduced in turn.