Classification Techniques In The Tax Administration

Data Mining is a very active research domain, which involves database, artificial intelligence, statistics and so on. It is the process of identifying valid, novel, potentially, useful and ultimately comprehensible knowledge from database that is used to make crucial business decisions. Classification of large database is an important Data Ming problem, and the application of classification and prediction technologies on tax collection has a bright prospect.Attribute-oriented induction method and classification technologies were particularly discussed in this paper. It is well known that different industries have different characters. The attribute-oriented induction method can be applied to find the characters of certain industry or the difference among different industries. By using this technology, we can classify the quality of tax collection and administration, and develop the criterion to evaluate it. This is why the State Administration of Taxation executed the categorized management on tax collection for quality improvement purpose.Decision tree classified algorithms has the advantages of high speed and efficiency, such as ID3 or C4.5 algorithms. However the efficiencies of these algorithms decrease obviously when the sample data size grows over a critical level. SLIQ, a fast and scalable decision tree classified algorithms, was developed to solve this problem. In this paper, we used this technology to produce a decision tree based on existing general taxpayer's information and predict their credit according to the output of the decision tree. This technology can reduce the range of check and increase the efficiency of our work.The SQIL algorithm was accomplished by using Java language. We also published it on our intranet to share the information.