Prototyping data warehouse systems

Abstract

The process of developing and evolving complex data warehouse systems is intrinsically exploratory in nature. Data warehouse projects are notoriously difficult to manage and many of them end in failure. One explanation for the high failure rate is the lack of understanding of the requirements and missing proof-of-concepts for a decision support for the knowledge workers of an organization. In this paper, we present a prototyping approach for data warehouse environments. We show how prototyping can support an incremental and iterative requirement development. We introduce different types of prototypes and illustrate their applicability for data warehouse systems. Finally, we show how to develop data warehouse prototypes with the BEDAWA tool, which supports a rapid development of prototypes by automatically generating the mass sample data for a data warehouse system.