Most of us, the BIDW Practitioners, would be used to the words ‘Kimball & Inmon’ in our daily lives. Each of us would be having our own views on the merits and demerits of adopting these different school of thought.
The Kimball approach focuses on the star schema based design and in some areas where the complexity is high even snow flake structures are also used. These are nothing but an extension of the star schema. Whereas, Inmon approach revolves around a relational, non redundant, granular, integrated design of detailed data. Datamarts are created based on the different lines of businesses based on this underlying data model.
Recently Bill Inmon had written a piece in the B-Eye-Network on ‘Kimball Vs Inmon‘ and why he thinks that Inmon model is the best. Areas of disagreement especially from the architectural perspective has been nicely put forward with due justifications.
However, given the fact that there still exists many Data Warehouses which had been designed using the Kimball approach with star schema in place and are also found serving the customers extensively, the discussion becomes even more interesting. Also, adding to the twist, quoting to Inmon himself, is declining of the invite from him to Kimball at least four times to debate and discuss on this in a open forum.
What is your take – Kimball or Inmon? Happy to listen to your comments.