Materialized View Selection in Data Warehousing

Everyday an enormous amount of data is retrieved and transmitted. To avoid the chaos, data mining and warehousing is used. A data warehouse is a set of materialized views defined over remote source relations. These materialized views must be maintained in response to actual relation updates in the remote sources. During the initial design and evolution of data warehouse, the designer has to handle problems of selecting views to materialize in the data warehouse. A design objective can be the minimization of a cost function or a constraint of available space. AND/OR DAG expression is used to represent alternative ways of evaluating multiple queries and views and sub expression sharing. Proposed approach can be used to optimize the views for the better performance of the data warehouse