Databases Reference
In-Depth Information
everyone uses the terms discussed here in exactly the same way. Sometimes
theBIlabelsareusedbecausetheyalignwiththenameofthesoftwaremodule
used to create the reports. Sometimes the terms are rooted in how your
organization has historically delivered data. It really does not matter whose
terms are used, but it is important for everyone on the project team and, it is
hoped, across the organization to use these terms in a consistent manner.
BI Building Blocks
When you use a BI solution you tap into several different layers. These different
layers are created to enable flexibility and the sharing of components across
many reports and applications. When using a BI solution, you generally are not
aware of these layers or how they work together. It is useful to discuss them
here because it will help the entire team to have more meaningful discussions
when refining requirements and developing the applications for the DW.
Data Content—Understanding What You Have
The foundation of any BI solution is the data itself. All of the data must have
useful business names and descriptions, rather than technical names used by
the database. There are two parts of defining the data content.
First, the data itself must be set up so that it can be used to create and navigate
through the BI reports and other display mechanisms. This is where the
Business Dimensional Model described in Chapter 7 is used again. Each of the
business attributes for each business dimension must be defined to the BI tool.
This entails linking the business data element to the physical data element in
the database. Additionally, the BI tool needs any relationships between each
of the attributes to be defined. This is how the BI tool ''knows'' what to drill
down to when asked. The labels that are set up here are what show up on
reports and as labels for setting parameters. This is setting up the BI tool to
present what is defined in the Business Dimensional Model as discussed in
Chapter 7.
Second, the actual business measures that appear as the body of the
reports must be defined. This work will use the Business Measures Work-
sheet described in Chapter 7. Think of this as setting up a library of all of the
business measures that will be used. These may be basic facts, key performance
indicators, or other business metrics. Each should have a descriptive name
and a business description, and each should define the default handling of
this measure for rolling the results up. This is also called the default aggregation
rule . For sales units, for example, the default is usually to perform a sum. For
an average price measure, the default aggregation rule should be to sum the
dollar sales divided by the sum of the units sold, rather than the average of
the average price itself.
Search WWH ::




Custom Search