Application Of Data Mining And Data Warehouse In The System Of Tax Payment Evaluation

With the development of tax revenue business, tax payment evaluation has become an important part of tax revenue management system in our country. However, tax payment evaluation system established in traditional technical method exposes many problems, affecting the accuracy and fairness of tax payment evaluation. To solve the aforesaid problems, this thesis applies the technology of data warehouse and data mining to introduce a set of system to intellectively select objects of tax payment evaluation in consideration of the need of tax payment evaluation business, in this way, quality of tax payment evaluation efficiency is improved.Firstly, the thesis analyzes the application of date warehouse technology in the tax payment evaluation, establishes corresponding data model, and improves the efficiency of tax payment evaluation in evaluation index calculation and data inquiry analysis.Secondly, this thesis studies each kind of data mining algorithms, especially deeply analyzes data mining classfication algorithms, building tax payment evaluation model selecting evaluation objects according to tax payment evaluation index. This model selects tax payment evaluation indexes based on analysis of correlation between attributes, builds two models based on selected indexes:algorithms of decision trees, algorithms of native bayes.The thesis evaluates the accuracy of two models, and analyzes two model's applications in various situations of tax payment evaluation businesses.