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

Silverston c08.tex V2 - 11/21/2008 3:12am Page 417

Level 2 Business Rules Pattern 417

Often people regard process modeling (where many business rules are

captured) as completely distinct from data modeling. Enterprises often capture

business rules independent of data models in process models, rule dictionaries,

metadata repositories, or rules engines. However, we feel that while these rules

may be captured in other types of models, you can achieve great advantages

by maintaining many business rules in the data model. Rules and data are

often tightly linked, and there are many situations where it is beneficial t o

maintain them in the same model. We believe data modelers should have

templates for how to capture data about rules in the data model, and use this

for the business rules they come across in the process of doing their analysis.

When business rules are maintained in a data model and then implemented

in a subsequent database design, they can then be dynamically changed. For

instance, if you use a data model to maintain all the various rules regarding

how a product is priced (for example , 2 percent discount when more than

100 are sold), then when that rule changes (for example, to 3 percent discount

when more than 100 are sold), you can often simply change the instances of

the data model (and the database that was based on the data model) without

changing the structure of the data model