Microsoft Office Tutorials and References
In Depth Information
FIguRE 4-28 A star schema has a fact table at the center and dimensions around it.
When you look at this picture, you can easily understand that a Customer bought a Product
with a specific Currency. The sale is pertinent to a specific Territory. Star schemas are easily
understandable by almost anybody at first glance.
As long as a data warehouse contains only star schemas created in accordance with the Kimball
methodology, it gains several other very useful characteristics:
■ There are no chained relationships. Because relationships exist only between facts and
dimensions, the maximum depth of a relationship is 1. This makes it easy to generate
relationships in PowerPivot because you always know that a dimension is related to a fact.
■ All information is highly denormalized. This means that you never need (apart from
some very specific cases) to use the RELATED function to load a column from a table
into another one.
■ If the model is well written, all descriptions in attributes are user-friendly. This means
that no columns contain NULLs and the user can slice the data by using attributes in
a very easy way.