A Novel Approach of Using Materialized Queries for Retrieving Results from Data Warehouse

Abstract

AbstractThe organization data warehouse stores historical records collected from various operational sources for management strategy making. In case of frequent management queries, generation of same results by repeated invocation to the data warehouse is relatively time consuming. In order to extract results from data warehouse, data-cubes and materialized-views are used. They acquire more processing, storage area and maintenance cost. The present study enhances result fetching of the queries which are frequent from a data warehouse by loading queries, results and some meta-data in a database stated as Materialized-Query-Database denoted as MQDB. When a query is given by the user, the MQDB is checked for determining any already existing query in the database. In case the query exists in MQDB and the stored results do not require any result updation then the results are simply fetched from the database. The approach of storing the results of the input query and thereafter simply fetching them when the same query is executed again reduces the query processing time substantially. The evaluation of the novel approach done by using the data warehouse on the central as well as on a remote cloud-server shows a noteworthy reduction in the time taken to retrieve results of the queries as compared to using the prevailing approaches. The approach is suitable to make use of past records in the data warehouse for management decision making.Keywords Data Warehouse Online analytical processing queriesMaterialized queriesFaster query resultCentral-serverCloud-server