Research Of OLAP And Data Mining Technology Based On Water Supply Data Cube Of Quantity And Charge

With the rapid development of information technology, especially the databasetechnology, the business system of enterprise has accumulated a large number ofvaluable data about the client and business. The traditional Online TransactionProcessing System (OLTP) is operation-oriented. It only can provide simple querying,inserting and deleting operation, and generate specific report. But with the databecoming huge, the traditional OLTP only obtain the surface layer data information.However, the connotative information and internal relation of the data cannot beobserved, which will waste the data resource. Because of that, people need anintegrated environment for the purpose of rapid querying, providing strategicinformation for analysis, discerning trends, and monitoring performance. With thedevelopment of data warehouse, online analytical processing and data miningtechnology, they provide an effective way in solving this problem.Firstly, this paper presents the research purpose of building the water supply datacube of quantity and charge. Meanwhile, it introduces the basic concepts of datawarehouse, online analytical processing and data mining. After that, the paper analysisthe business data of the Water Supply Company and designs the model of data cubebased on the requirement. The model is mainly composed of three areas: conceptmodel design, logic model design and physical model design.Secondly, we extract, transform and load data from the database to fact table anddimension tables. According to the Analytic Workspace Manager of Oracle, weestablish the data cube of quantity and charge and OLAP on it. Finally, the paper discusses the cluster data mining on the data summarized bythe cube. We use the algorithm named k-prototypes which can both handle nominaland numeric attributes. The algorithm is achieved by the MATLAB program and itclusters the customer to several kinds successfully.