The research of a new tax assessment model

Tax assessment is an important and complex task in tax administration. The commonly tax assessment methods only use tax-related data for classification, which can not meet the needs of the actual tax administration. Tax assessment usually takes decision tree algorithms as a method of classification. General decision tree algorithms take information gain as classified target, the training sample collections or the class tables request resident memory. The SPRINT algorithm uses Gini index to select attributes and eliminates all memories limit. The SPRINT algorithm and comprehensive fuzzy evaluation can be combined to generate a new tax assessment model. With this model a sequence can be generated by evaluating the tax-related data and the basic information of taxpayers. This model can solve the issues that the taxpayer´s order of tax assessment which usually can not be scientifically determined. Experiments show that the decision tree algorithm can improve the accuracy of classification. The tax assessment model based on comprehensive fuzzy evaluation can evaluate the taxpayers, with this model a sequence can be generated for further checking. This tax assessment model is valuable for tax inspection.