ABSTRACT

Data warehouse is set up for the benefits of business analysts and executives across all functional areas. The primary goal of data warehouse is to free the information locked up in the operational database so that decision makers and business analyst can make queries, analysis and planning regardless of the data changes in operational database. Data analysis in a large database requires some innovative techniques; there should be some new methods and procedures to look at the data with different angles. While executing adhoc queries over a large database degrade the query performance at considerable level because query will have to go through from large volume of data after making lot of joins. Moreover, as the number of queries is large, therefore, in certain cases there is reasonable probability that same query submitted by the one or multiple users at different times. Each time when query is executed, all the data of warehouse is analyzed to generate the result of that query. In this paper we purpose an ideal approach which definitely minimizes response time and improves the efficiency of data warehouse overall, particularly when data warehouse is updated at regular interval. This approach has been validated by Formal Method.