Selection And Adjustion Of Materialized View In Data Warehouse

A data warehouse is, by definition, a subject-oriented, integrated collection of data which support managers to make correct decision. It enables complex queries on data warehouses, which contains large amounts of data. However, How to manage so much data efficiently and provide high efficient query is one of the problems faced by data warehouse. And materialized view is a important method to solve that, but it needs extra system space to store and also needs to be maintained at the price of system immolation. Therefore, The selection of materialized view is an important research subject. The static selection algorithm of traditional materialized view is based on the assumption that probability distribution of queries has been provided by the user, or these queries in the integrated data are uniformly distributed. But these assumptions always can not be realized in practical use. So a view selection algorithm with the capacity of both static selection and dynamical adjustment is quite practical and researchable.This paper lucubrate the selection of materialized view from static and dynamic aspects and presents a new materialized view selection and adjustment algorithm --VSAA (Views Selection and Adjustment Algorithm) which based on the SPJ (Select-project-join) view assumption data warehouse model .This algorithm take MVPP(multi-view processing plan) as the search space and integrately consider the factors of storage space, view maintenance and query capability. After having researched the VSAA theoretical mode from mathematic model and key parameter aspects, this paper put forward DVMV (Dynamic Virtual Materialized Views) algorithm as a supplementary of VSAA algorithm specific to the real-time defect of VSAA. And take the theoretical analysis as guidance to realize the VSAA algorithm as well as contrastive analysis to other classics algorithms. Finally, I use tpc-h benchmark data to generate 1G data into Oracle database as the experimental data, and proved the effectiveness and superiority of the algorithm VSAA through theories and experiments.