Atualização eficiente de visões materializadas em data warehousing.

Abstract

A Data Warehouse (DW) is a great data storage toll and support to the management decisions that are "fed" by diverse information sources that are distributed. These constitute the operational systems or production systems o f the organizations. One o f the important characteristics of a DW is that i t is temporal, wich makes it o f high volume. A typical report from a DW, called OLAP (Online Analytical Processing) report presents temporal series o f summarized information using different grouping criteria (business dimensions). I n order to maintain the costs of OLAP acceptable, an approach called Materialised Views was adopted. This approach store calculated summaries physically in the DW. By using these materialized views, the managing software is able to process the OLAP consultation with a good or acceptable performance. Materialized views are the central topic o f this research, mainly, the problem o f updating. I n regular intervals the DW, with the materialized views included, need to be uptaded so that the changes in fonts can be incorporated. This can be a costly process, requiring great effort in research, in order to improve the performance of updating algorithms o f the DWs. I n this context, we present an algorithm for the efficient updating of materialized views in DWs.