Research On Key Technology Of Social Security Audit Expert System

Social insurance is an important part of social security system,which have great significance to solve communitarian's worries behind,to resolve social contradictions,and to improve sustainable development of the economic and society,which is also a major event related to the people's livelihood.With the increase of audit scope and more complicated audit item,as auditing officers have different audit level,how to improve work efficiency and make use of audit experts'successful experiences is a outstanding issue in audit field.With the development of computer AI and the extensive use of the social insurance computer audit system,social insurance audit expert system is a effective way to solve the above problems.In this thesis,study the critical technology of expert system and the operational audit method of endowment insurance,on that basis to research and discuss the techonology of social insurance audit expert system.First of all to introduce the expert system information and knowledge representation,analyse and compare case-based reasoning and rule inference and study the integration of the two,then introduce the operational audit of endowment insurance.Use the producion rule to conduct a knowledge representation for endowment insurance treatment payment of the audit,individual contribution amount of the audit,.The individual contribution base for the audit and the individual contribution percentage of the audit,and use xml to represent the proportion of audit fees case, building a knowledge base of social security audits,to research the knowledge acquisition of social security,and design the redundant processing algorithms of the old-age insurance benefits to pay for the audit rules in knowledge base.To achieve the rule inference of endowment insurance treatment payment of the audit and the case-based reasoning of the proportion of audit fees.Use the rule which is relational database storing the audit pension insurance benefits paid,to relize the rule inference of endowment insurance treatment payment of the audit.In case-based reasoning, using the DTD defines the proportion of the audit fee structure of case, by CBR engine to retrieve the proportion of audit fees case from the case-base, and then use the AHP method to calculate the weight of the characteristics of properties of the proportion of audit fees case, and conduct the conformance testing.At last use neighbor method to reason the proportion of audit fees case and case-case study of self-learning. And design the weight calculation and consistency of detection algorithms.