Application Of Data Warehouse And Data Mining Technology In Tax Administration System

The database techniques are the key parts of decision support. Mature techniques of database is also foundation of DW(data warehouse) techniques that bring up, but it is decision that DW and DM(data mining) to support the new techniques, it will become core techniques of the integral information support system.According to above various techniques, this thesis expatiate application techniques of data warehouse , how to establish data warehouse of the tax administration system, and analyze multi-dimensional data model, metadata, the system construction, techniques of the data warehouse requests. Finally, we construct the data warehouse of the tax administration system. Because the current business process system contain various data source in tax administration system, this thesis comprehensively gives to the strategy of realizing concerning to data cleaning, integration, transformation and data load in data preprocessing. And carried out the OLAP the operation in the fact form that establish. Through these operationses realize data warehouse application in tax administration system. Studying the logic & the physics of the data warehouse designs , including the data model analysis, multi-dimensional database design, the data warehouse parallel processing, partition etc. On the basis of overall situation , we put forward a kind of realize method , which from the data mart to the data warehouse in tax administration system. Finally, the thesis analyzes index techniques of the data warehouse, method to import and export and problem with the safety & control.In the research of DM technique, the paper discuss the function, classification, task , methods , technique, tools and model of data mining. Put forward the basic procedure and main steps about application of DM in tax administration system. Analyzing the data mining primitives, query languages, and system architectures. We consider the comprehension and analyze of algorithms for association rules ..provide an overview of association rule research and improved algorithms of Apriori. Considering revenue business, we calculate support threshold of itemset and confidence threshold of rule , realize application and prediction by association rule mining in the audit department of tax system . By algorithms we analyse various illegal possibility, find out some implicit regulation among of factors , and guide current revenue work in audit.We research the basic concept and function of DSS(Decision Support System), and analyze application circumstance in the domestic and international tax administration system. On the basis of revenue information data and administration law of tax in State Administration of Taxation of China, we mainly design and analyze the topic model, andABSTRACTput forward a frame of DSS. Through function analyzes of the tax administration' s DSS, we put forward the four type problems of DSS. Finally, we are acquainted with the basic concept and tools of OracleQi OLAP , and realize statistical query of data by web .and PB program.Because currently domestic tax administration system haven't complete application project of warehouse and data mining techniques, this paper put forward a full range of application project. These techniques have advanced and practicability in domestic tax administration system , and instruct effectively revenue work.In particular, we make three primarily contributions. First, we design and establish DW of the tax administration system. Second, we realize application and prediction by association rule mining in the audit department of tax system. Finally, we systematically design the topic model, put forward a frame of DSS, and realize statistical query of multi-dimensional data by ASP and PB program.