The Application Of Cluster Analysis Model In The Tax Assessment

Data Mining is a potential and useful rule which came from large amounts of data using mathematical, statistical, artificial intelligence and neural network to identify. In recent years, it has received the widespread attention in the field of domestic and international numerous Data Mining, clustering analysis is an important research topic in the data mining, it divided a large amout of unknown data object into multiple categories according to the data of the internal similarity, data objects is a mathematics method which is bigger within the category similarity and smaller between categories and its application field is very wide. In recent years, with the rapided development of socialist market economy, we promoted it vigorously into specialization, informationization, tax management, optimize the tax service, guard against tax analysis, which can be used to design the index system to extract the personality characteristics, using the cluster analysis technology carries on the comprehensive analysis and interpretation results to find out the abnormal tax based to complete tax assessment work better, so the search and application of clustering analysis is particularly important in tax assessment.Clustering analysis algorithm is introduced on the GuiYang’s "five one" local tax bureau tax professional management platform of tax evaluation module in the specific application in detail in this paper. By understanding the basic concept of clustering analysis, a preliminary study of the clustering analysis of each algorithm, and K-means algorithm in clustering analysis and system clustering method research, and through the use of statistical software SPSS Statistics to related data clustering analysis and interpretation in the actual work environment, finally used in the tax evaluation module. NET technology integrated the analysis function of SPSS mathematical model which through the use of third-party DevExpress’s controlled implementation of the front-end interface display, the content mainly includes:Model1:The clustering analysis of pointing regional tax burden and GDP growth;Model2:The clustering analysis of main tax revenue contribution;Model3:The clustering analysis of tax revenue and sales income growth; Model4:The clustering analysis of enterprise income tax and profit margin;Model5:The clustering analysis of key tax sources door enterprise income tax and profit margin;Model6:The cluster analysis of key tax sources household tax revenue and sales income growth.Based on the above six model for clustering analysis and interpretation, it betterly interpretated the relationship between GuiYang’s local taxation and the local economy, effectively help tax staff distinguish abnormal user better, now the system has been used in GuiYang’s local tax bureau which have achieved good results.