Research On Tax Assessment Of Automobile Sales Industry In J City

In the era of big data,China’s tax modernization is inseparable from artificial intelligence and data mining.The tax application of big data is particularly important in improving the quality and level of tax collection and management in China.We will integrate big data into tax assessment and constantly optimize our tax assessment system.At present,there are still many problems in our country’s tax assessment work,such as: the accuracy of tax assessment model to identify risk taxpayers is not high,the efficiency of assessment work is low,the quality of assessment results is poor,the coverage of assessment is small,the subjectivity of assessment personnel is too strong to affect the fairness of tax,the assessment results can not form a rapid response case,etc.There are many reasons for the problems,the most important of which is that most tax authorities do not apply big data enough and lack effective processing methods.The core of tax assessment is risk identification.How to identify risk taxpayers quickly,efficiently and accurately has become the long cherished wish of the grass-roots tax authorities,and it also solves the trouble of enterprises being disturbed without risk.Under the existing evaluation model,the tax authorities inevitably have risk-free evaluation and problem-free inspection,and the evaluation time is long,which has an unnecessary impact on the production and operation of enterprises.The risk identification of big data + can make decision and judgment quickly according to the historical data of deep learning.Random forest has the advantages of fast training speed,strong generalization ability,high accuracy and simple implementation.This paper takes the automobile sales industry of J City as the research object,collects the tax report data of enterprises in 2017,selects 14 general tax assessment indexes of four categories related to the industry,namely,the enterprise tax burden rate,the enterprise profitability,the enterprise operation ability and the enterprise debt paying ability,establishes the tax assessment model of automobile sales industry of J City by adopting the method of forest classification,and points out the tax assessment In order to improve the prediction accuracy of the tax assessment model,the importance of the standard is sorted,the tax assessment model is simplified and improved.It makes the model more concise and efficient,easy to promote in the basic tax authorities,and promotes risk-free and zero disturbance enterprises.In the era of big data,this paper puts forward countermeasures and safeguard measures on how to mine data and how to use artificial intelligence for tax assessment.For the first time,this paper combines the random forest classifier with tax assessment,and establishes a tax assessment model of automobile sales industry in J city.This model is simple and practical,with high accuracy,fast response speed,ranking the importance of features.It can set up a scientific industry index system,shorten the time for tax authorities to find taxpayers’ risks,improve the accuracy of screening risk taxpayers,reduce unnecessary interference to risk-free taxpayers,and create a good tax business environment.It is hoped that this study can provide some reference for the grassroots tax assessors to use simple and efficient tax assessment model(random forest classification technology)for tax risk management,and provide some new ideas for the intelligent application of random forest classification in China’s tax revenue.