Research And Implementation Of Materialized View And ETL In Data Warehouse

Data Warehouse is a data set with characteristic of subject oriented, integrated, non-volatile, time variant. Data Warehouse used to find crucial important business information for corporations by analyzing history data. It can help managers to make decision correctly and quickly. With the rapid development of information business, Data Warehouse would be a research hotspot. Materialized view is a crucial technology to improve the performance of Data warehouse. It is a research hotspot to maintain materialized view. The performance of ETL(Extract,Transform,Load) determine the success of building Data Warehouse.This dissertation makes extensive and deep research on crucial technology of Data Warehouse. It introduces the architecture of Data Warehouse and further discusses the technology of ETL and maintenance of materialized view. The main content of the dissertation is summarized as follows:1. This dissertation review recent research in Data Warehouse and materialized view. We also elucidate the research hotspot and existing problems of Data Warehouse. Then we describe the architecture of Data Warehouse, the technology of ETL and materialized view in detail.2. An improved algorithm to reduce storage space for auxiliary view of multi-materialized view is proposed. The new algorithm find appropriate view to combine together to reduce spaces effectively.3. Design a system for ETL using metadata. This tool implement ETL with the perspective of source data and target data that inducted by metadata.4. Develop a test system of ETL. We illustrate each module in ETL execution order.