| Popis: |
Data warehouse (DW) has been widely recognized as an effective solution for integrating diverse information systems. In addition, DW supplies a platform for business intelligence applications which aims to improve the process of decision making. At execution level, after the first load, DW is regularly populated with fresh data by ETL (Extraction, Transformation and Loading) processes. In real setting, DW refreshment is an event managed by DW administrators. Still, non freshness of data is a drawback of classical DW system. In this paper, we suggest a solution for this issue by presenting architecture of near real time DW. We visit DW system layers and we show how they should cooperate. Finally, we define the scope and the challenge of each layer simultaneously with our suggestion to fill its mission |
