Dynamic management of materialized views in real-time data warehouses

The update frequency for traditional data warehouses cannot meet the real-time data analysis. To alleviate this problem, the real-time data warehouse (RTDW) technology has emerged. A RTDW allows decision makers to access and analyze very recent data as fast as possible in order to support real-time decision processes. This analysis process is performed using complex queries with multiple joins and aggregation operations on large data tables. Despite the complexity of their queries, decision makers want their requests to be processed rapidly. One among the optimization techniques used to respect such a time constraint is materialized views. The purpose of this work is to better manage materialized views and to deal with view selection as well as view maintenance problems in RTDW. In this paper, first we propose a dynamic selection of materialized views algorithm (DynaSeV) which selects views from results of incoming queries under the execution time and the storage space constraints of the system. Secondly, we suggest a new update policy that dynamically determines the method to maintain materialized views according to their access frequency and the system workload.