The degree of corruption or anti-corruption risk is an vital standard of political civilization. Scientifically evaluating, predicting and preventing the risk of corruption are the prerequisites for effectively curbing corruption and building wise politics. It is also one of the hot and difficult problems in the current study of how to build a wise government. China is a unitary state, and the risk assessment of regional independent government is of universal significance. However, similar studies are hardly found in the database of CNKI, VIP and Elsevier. To scientifically evaluate the anti-corruption risk of regional government, we cannot take a traditional simple qualitative assessment based on a single data with a large limitation. So, on the basis of theory reservoir, expert interview and questionnaire survey, in accordance with the construction of a system of indexes, comprehensive evaluation, empirical/predictive analysis. policy recommendations and other major steps, we carried out a combination research of qualitative and quantitative methods research. In the empirical region selection, considering the data source facilitating, This study take Jiangxi province as an example, where several provincial and ministerial level cadres for disciplinary violations are investigated and after the Eighteenth National Congress of the CPC, especially Su Rong, the former province secretary caused some negative impact in the whole country. Therefore, the choice of Jiangxi Province for the empirical research on risk assessment of anti-corruption risk, is of practical significance. Our main work is as follows:I. Concept definition and system analysis.The concepts of anti-corruption risk and regional anti-corruption risk were defined using the methods of system analysis. Through, HSM and SSM analyses of regional anti-corruption risk evaluation, the 3D regional anti-corruption risk evaluation and soft system analysis model were set up.II. Construction of anti-corruption risk assessment index system.Based on the regional anti-corruption risk formation mechanism and influencing factors analysis, five objective factors, i.e. "opportunities for corruption, incentive and protection, supervision, punishment and regional environment" and four subjective factors, i.e. " degree of attention, clean degree, treatment effect and public confidence " are put forward to building the preliminary draft of regional anti-corruption risk evaluation index system. More than twenty experts in the Tsinghua University, Beijing University of Aeronautics and Astronautics, and the Chinese Academy of social sciences were chosen for consultation questionnaire, using reliability, validity and Ridit analysis method to carry on the statistical sifting, and thereby determining the two level indexes, 9 second-class indexes and forty-eight tertiary indicators for regional anti-corruption risk evaluation index system..III. Carrying out regional anti-corruption risk assessment and empirical study.Compared with only one evaluation method, the comprehensive evaluation of various methods is more objective and fair, and the result is more closer to the reality. Firstly, based on the MATLAB software, the improved hierarchy analysis method, with cluster analysis, power method, similar coefficient weighting method to determine the index weight, and perform the empirical assessment of Jiangxi province from 2010 to 2014, a total of 20 quarters independent risk situation. Secondly, based on the level of analysis, the fuzzy analytic hierarchy model and grey analytic hierarchy model of Jiangxi Province between 2010 and 2014 of 20 quarters are studied to evaluate each anti-corruption risk situation, and fortunately, three kinds of evaluation methods came to the similar results. Subjective and objective indicators trend is similar, assessment results are in accordance with the actual situati which shows that the regional anti-corruption risk assessments have certain scientific property and feasibility.Empirical results also show that the assessment results of anti-corruption risk in Jiangxi province from 2010 to 2014 show the trend of from bad to good, and. overall risk is in the lower level of risk. However, the "supervision and control" "regional environment" and other indicators of higher risk assessment, objective risk assessment index value is higher than the subjective risk assessment index,these all need to bring to the attention. Furthermore, the anti-corruption risks in Jiangxi province fall in a declining trend after the Eighteenth National Congress of the CPC; the change of the central and local political cycle of Jiangxi Province have greater impact on the anti-corruption risk; Leadership changes in the short term does not have a significant impact on the risk of corruption in Jiangxi province.IV. Regional anti-corruption risk prediction and empirical analysis.According to the comprehensive evaluation results, the gray GM(1, 1) model, BP neural network model and SVM model were used to predict the risk of eight quarters of 2015-2016 in Jiangxi province. The predicted results given by the three methods agreed well in Jiangxi province from the first quarter of 2015 to the fourth quarter of 2016, and the risks decline. From the prediction accuracy, SVM is superior to BP neural network and grey GM(1, 1) model.V. Factor analysis, system based model analysis and regional anti-corruption risk prevention and control strategy.The nine risk factors of two level indicators were analyzed, and four main factors were extracted to establish the risk prevention and control model of regional anti-corruption risk, i.e. "drinking poison to end thirst", "growth limit", "target erosion" and "the rich getting richer" fundamental mode. Then, the guiding ideology and specific measures for the prevention and control anti-corruption risks of regional government were put forward, and the countermeasures against the anti-corruption risk of Jiangxi province were given.The work in this dissertation is derived from the project of "report on combating corruption and upholding integrity in china NO.5" sponsored by the Chinese Academy of Social Sciences Research Center, and some research results were published in the journal of "Statistics and Decision", "Jiangxi Social Science", "Truth" and others. |