Dynamic Materialized Views Selection with Scalability in Data Warehouse

A data warehouse usually needs to process a large amount of data and gets concise results to answer queries committed by users.Because of this point,materialized views are of unprecedented importance in data warehouses.The current materialized views selection approaches are static,which greatly disobey dynamic nature of OLAP and DSS.In this paper,a scalable dynamic materialized views selection is proposed,which divides a whole MVS into three phases to reduce the complexity of the MVS problem,the effectivity of materialized views is improved as well.Through dynamically adjusting materialized view,materialized views can adapt on-the-fly to incoming query pattern.Complexity analysis proves scalability.Empirical evidence demonstrates that our dynamic adjusting algorithm is effective,and that our approach is self-adaptive.