The Study Of Government Agent Evaluation Based On Rough-set Theory

Under the background of rapid development of the current economy and society,in order to enhance project management efficiency,the management mode for nonprofit government projects has gradually developed from self-management by contractors and contract-based management to management by construction headquarters and project artificial person responsibility-based management.However these new management modes are still incapable of completely solving problems in government project management;a lot of government projects are suffering low investment efficiency and relatively low management level,etc.In order to improve the construction and management efficiency of government projects and promote the extension and development of the "agent-construction system" for government projects,the State Council has issued a number of rules to reform the government investment system and the agent-construction system has been applied in some cities for trial.In view of the propagation and application of the agent-construction system in China,a rough set-based government agent indicator system is established and applied in agent evaluation.The main contents of the paper are as follows:Firstly,the paper describes agent system models and relevant concepts and then studies the credit risks of agents and indicates that such risks are mainly caused by information asymmetry,incompleteness of contracts and sudden risks.For agent risks come from legal system,government regulation and incentive mechanism,they are inevitable in project agent-construction.Secondly,a sough set-based agent indicator system is set up.Upon introduction of the construction principle of the indicator system and rough-set related theories,indicator initial selection,duplicates removal and rough-set indicator method screening are designed.When constructing the three-grade indicator evaluation system according to the three-stage construction method,the meaning of each three-grade indicator is described.Thirdly,an AHP-entropy evaluation model is discussed.Upon instruction of the AHP method and entropy weight concepts,the entropy weight concepts are applied to establish an AHP-entropy evaluation model which performs better in overcoming the problem of more subjective factors in the AHP method and obtaining a more effective weight.Finally,the method constructed in this paper is applied to actual agent screening to verify its effectiveness.This paper shows that the rough set theory can better remove subjective factors in the AHP method and the rough-set-based agent indicator system can be favorably applied to agent screening,providing a new method for agent screening.