Application Of Data Mining Project In Statistical System Of Review Cycle

In order to give the Bureau of XX to establish more efficient and scientific review of businesses operation and management system, continue to improve review efficiency and review quality, to create a more flexible, convenient and efficient review to provide a scientific basis; in order to enhance the review of monitoring, evaluation, statistics, analysis ofwork automation, the level of information; statistics in order to expand the scope of online services, and enhance the transparency of business processes, the review cycle in order to dig deeper into the review process, findings and review of quality of information and provide decision support for the review of business management and strategy formulation; system construction projects, this project is the background were analyzed.This article needs analysis of the project, the design of the overall architecture design,data warehouse and data mart models, data warehouse and data mart design and development of the whole process and report to achieve a detailed description and analysis. Related technologies used in this paper include the ERwin modeling techniques, data mining, data warehouse and OLAP technology, cognos report development technology. This article addresses four technical difficulties:the integration of application systems integration and expansion of the processing of heterogeneous databases, data standardization and data quality control, multi-dimensional analysis.The joint efforts of the project team members, the project completed on schedule, but there are still some problems found during the development process, such as data import efficiency, a prominent feature of this system is the huge amounts of data, even though we write a stored procedure and call a stored procedure has been the greatest extent possible optimization, data import too long, or do we need to be resolved and discuss the problem. There cognos development statements demonstrate a lot of advantages, but there are some shortcomings, such as in mapping, there is no tabulation of more intelligent, according to the needs of interception. These issues will be continuous learning and gradually perfect solution.