The Data Model Resource Book, Volume 3: Universal Patterns for Data Modeling

Silverston c09.tex V2 - 11/21/2008 3:16am Page 492

492 Chapter 9 ■ Using the Patterns

Synopsis

In this section, you saw the creation of an application data model to support the

needs of Sands Distribution in relation to its customer information application.

You saw how the data team evolved the prototype model seen in Figure 9-2

into a full-blown application data model. They did this by replacing some of

the specific patterns (level 1 and level 2) with more flexible patterns (level 2

and level 3).

The data team decided to balance accommodating specific needs with a

flexible style of modeling. Meeting specific needs allowed the data model to

be customer–tailored to the application. Using a flexible style of modeling

allowed new data requirements to be added more easily. This meant that they

used different levels of patterns depending on current and any anticipated

future requirements. For example, they added the PARTY and PARTY ROLE

level 3 concept to allow new declarative roles to be added more easily in the

future and also to facilitate capturing common PERSON and ORGANIZATION

information only once, regardless of the number of roles that a party may

play. On the other hand, they maintained more specific classifications of

CUSTOMER because the business said that these were the only way that the

customer application would classify customer data.

Some data modelers may object to using different styles and