Title:
Automatic Identification of Materialized Views Maintenance Scenarios
in a Data Warehouse
Abstract: The maintenance of materialized views in a data
warehouse depends on the choice of the scenario. The scenario must give a
response to when and why maintain the materialized views. The most maintenance
timings used in the literature are immediate, periodic or on-demand maintenance.
However the manners are incremental or views recomputing. In this paper,
we propose an algorithm that determines the appropriate moments and manners
to propagate updates from sources to materialized views. We propose our evaluation
criteria based in the current and the required data freshness levels. The
aim is to release the administrator from the manual research of maintenance
scenarios and to provide the adequate ones.
Authors: Ines
Ben Tekaya, Ali Ben Ammar, and Abdelaziz Abdellatif