Data Mining Research On Sampling Of Tax-checking

Now, Data Mining (DM) research is developing very quickly, the stage of information development in the Tax Administration System(TAS) is turned to the stage of data integration. However, the projects related to Data Mining is lack. Therefore, the research and application of DM in TAS has the important theories meaning with the actuality worth. There are three main research objects in this paper: First, Introducing the background knowledge of the tax administration business, and research and application present of the technique of DM at the tax administration system. Analyzing the applied foreground of DM in TAS. Second, Discussing the Association Rules analysis method, analysis the number-association method based Apriori. And we can adopt this method ,finding the appropriate indicator of tax-checking that hasing judgment ability appropriately and association weakly. Third, Make use of clustering method, on base of the indicator chosen, find the character of honest ratepayer and non-honest ratepayer.