Researchof Data Warehousing Modeling Basedon BW Platform

Since the concept of datawarehouse come into usage, it has been widely used in the dataprocessing part of the traditional fields of finance, assurance and telecom. With thecompetition of the industry and science, datawarehousebecome a power of marketing andselling and furthermore it becomes an important way to acknowledge customer's requirementand gain their information. OLAP is objective to the process and analysis online data, itmainly used to support complicated query which lready been defined clearly. Currently thereare three main patterns for the storage of datawarehouse: ROLAP, MOLAP and thecombination of both. The third pattern has been widely used in the domestic but has someshortage like low efficiency of data storage; the controlling of data access is risk.Based on the above background, this article will based on the theory of data warehousing,discuss the technology of data warehouse and its related fields. This project will combine thereal business case with the modeling technology of data warehouse, optimization of modelingdata warehouse, optimization of master data maintenance process, in order to improve andoptimize current modeling tool of data warehouse. Detailed information including:First, this article will introduce the relevant concept and its application usage. On theother hand, it will analyze and graphically explain the date warehouse from the views of theconceptual, logical and physical model.The modeling in this system is based on BW datawarehouse platform which is different from ordinary database; BW data warehouse platformcould enable multi-dimensional query and data mining.Second, based on the BW data warehouse platform, according to the modeling habit andstyle of key user, try to optimize the modeling process of traditional BW data warehouseplatform, optimize user experience and the model which user create in the BW platform, BWoptimizealgorithm is designed based on the dynamic modeling concept, it solved theperformance issue caused by transaction data increased during daily usage.Finally, explain the detailed system realization process, including system physical,functional and some interface implementation.Through the design and technology, this system could help end user (who doesn't haveany BW modeling experience) create model successfully easily. End user could also maintainthe master data and transaction data efficiently and easily. This system could optimize the user created model based on the data volume to improve the model performance dynamically.Based on the above theory and technology, this system has strong practical usage andtheoretical significance.