东北大学数据挖掘第4章.pptVIP

Why data warehouse? Businesses have a lot of data Data is available but in different formats and locations Decision makers want to access information virtually on one single site This access needs to be fast regardless of the size of the data What is data warehouse? (I) Data warehouse is a subject-oriented, integrated, time-variant, and nonvolatile collection of data in support of management’s decision making process W. H. Inmon ’s definition (1992) 数据仓库是支持管理决策过程的、面向主题的、集成的、长期的、稳定的数据集合 Data warehousing: the process of constructing and using data warehouse What is data warehouse? (II) Why subject-oriented? a data warehouse is organized around major subjects Why integrated? a data warehouse is usually constructed by integrating multiple heterogeneous data sources Why time-variant? data are stored to provide information from a historical perspective (5-10 years in data warehouse vs. 30-90 days in operational databases) Why nonvolatile? update of data does not occur frequently in a data warehouse (without transaction processing, recovery and concurrency control mechanisms) Multi-tiered data warehouse architecture Data warehouse vs. Heterogeneous DBMS Traditional heterogeneous DBMS Build wrappers/mediators on top of heterogeneous databases Query driven approach when a query is posed, a meta-dictionary is used to translate the query into queries appropriate for individual heterogeneous sites involved, and the results are integrated into a global answer set require complex information filtering, compete for local resources Data warehouse Information from heterogeneous sources is integrated in advance and stored in warehouses Update driven approach query and analysis are performed directly support complex multidimensional queries, does not compete for local resources Data warehouse vs. Operational DBMS Operational DBMS On-line transaction processing (OLTP) Day-to-day operations