Data warehouse design for ease of data transformation

Classical dimensional modeling aims for simple data warehouse designs in order to make the job of the end user and the DBMS easier. This paper describes a different approach aimed at simplifying data extraction, transformation, and loading tasks. Our design approach avoids and isolates changes to source data tables by surrounding them with helper-views. Each helper-view specializes in a single type of transformational task. The user is presented with a final view, which integrates the information from the source tables and the helper-views. We used this design approach in a data warehouse for analyzing half a million supplier bids submitted during Business-to-Business online procurement auctions.