Representative sample data for data warehouse environments

Abstract

The lack of sample data for data warehouse or OLAP systems usually makes it difficult for enterprises to evaluate, demonstrate or benchmark these systems. However, the generation of representative sample data for data warehouses is a challenging and complex task. Difficulties often arise in producing familiar, complete and consistent sample data on any scale. Producing sample data manually often causes problems that can be avoided by an automatic generation tool that produces consistent and statistically plausible data. In this paper, we determine requirements for sample data generation, and introduce the BEDAWA tool, a sample data generation tool, designed and implemented in a 3-tier CORBA architecture. A short discussion on the sample data generating results proves the usability of the tool.