Application Research Of Classification Method On Evaluation Ratepaying Credit Grade

Revenue is the main source of a country's finance, and the important tool to adjust its economic development. Revenue credit is the important part of society credit system, and sincere ratepaying represented social credit situation in some degree. Nowadays, our revenue credit system is under the building step, ratepaying credit grade evaluation is still finished by hand or simulated manual work by computer. Taxations development provide large mount of operation data for ratepaying credit grade evaluation, How to apply data mining technology to discovering knowledge, realizing the automatization and intelligentization, is one of the difficult points of taxation informatization. Upon this problem, this dissertation made research with data mining classification technology, and its main job are:(1) Analyzed and summarized the aim, content, standard, inspirit, monitor and current situation of taxation system ratepaying credit grade evaluation, developed a method that using decision tree to solve the problem of ratepaying credit grade evaluation by hand.(2) Discussed the aim, mission, technology and method of data preparing, and made data choose, data pre process and data covert to ratepaying credit grade evaluation data according to the data prepares process, form the data mining prepare data.(3) During the application research process, for the low efficiency of C4.5 algorithms continuous attributes discretization on real ratepaying credit grade evaluation, this dissertation developed a experience value based window split ratepaying credit grade evaluation algorithm, and the application showed a good performance, which means it is suitable for taxation system.(4) Applied the improved C4.5 algorithm to ratepaying credit grade evaluation, realized the auto evaluation of ratepaying credit grade, improved the efficiency of ratepaying credit grade evaluation work, reduced the disturbing of manual work, which made the whole process appears to be more justice and justness.(5) Based on the upwards research, prototype experiment system was build, and proved the correctness and efficiency of algorithm by theory and experiment.