The Research And Implementation Of Invoice Integrated Service Analysis And Decision System Of Local Tax Bureau Based On DW And DM

With the rapid development in Data Warehouse, OLAP and Data Mining techniques, great progress has been achieved in Decision Support System (DSS). An integrated DSS is mainly based on a Data Warehouse platform, with the use of OLAP and Data Mining tools. DSS is a solution rather than a product. An accepted rule in the realization of DSS is to layout a system structure, implement the system incrementally, benefit from the very beginning, and improve the design over time.hi this thesis, the concepts, implementation and development of Data Warehouse and Data mining are introduced briefly. The focus of the thesis then centers on applying these techniques to the development of an Invoice Integrated Service Analysis and Decision System of Local Tax Bureau. As it is well known, there has been great demand in meeting the new challenges that governmental revenue departments are facing: better client services, strengthened tax administration, improved resource management and the requiremet in evaluating new tax rules and policies. The proposed system utilizes OLAP and Data Mining techniques to analyze data in Invoice Integrated Service Data Warehouse on a multi-dimension and multi-level approach. It provides a solution for local tax bureau management to perform analysis and decision-making about Integrated Service of InvoiceA framework has been built for the solution of Invoice Integrated Service Analysis and Decision System of Local Tax Bureau. Modules within this framework have been generated. The design of the data warehouse uses the DB-BW-DMs (Independent Data Marts) structure. Data pre-processing is dealt with the DTS tools. For OLAP-analyzing and data mining on multidimensional data, the Multidimensional Expressions MDX, and Microsoft ADO MD data object techniques are used. Report processing and generation are realized by using Microsoft Excel 2000, which is one of the client application programs that support PivotTable Service. Data Mining Module realizes creating and training data mining model programmatically by using DSO (Decision Support Object). And results of analysis are showed by visualized mode. Examples have been given through the thesis to demonstrate the concepts and support the design. Conclusions have been provided at the end of the thesis to summarize the current work, and to point out future research directions.