Business Rules: Capturing the Most Elusive Information Asset - High-Performance Web Databases: Design, Development, and Deployment - page 117

rules and business-specific rules should be required on completion of
detailed design. Business rules should be well understood before the
management-level or detailed data model is considered complete. Data
administration should effectively enforce the submission of business
rules either from business people or from the information systems staff to
the business rule administrator before the creation of physical database
structures. Including business rules as a separate deliverable in the SDLC
at this time gives data administration greater leverage to ensure the rules
are formally captured and agreed on before application implementation.
The business rule administrator should round out the rules at critical
points in the SDLC, ensuring business rules match the data model. When
reviewing a management-level data model, common business rule defini-
tions should exist for each entity and relationship. In reviewing a detailed
data model, entity identifier rules for each entity, referential integrity
rules for every relationship, and a rule for each supertype data structure
should exist.
Presentation of Data Models
Data administration can reinforce the importance of formally document-
ing business rules by presenting them to the audience along with the data
model at a data model review. This is an excellent opportunity to uncover
discrepancies in business rules because the audience will likely be cross-
functional. Reviewing each business rule in a joint database review meeting
streamlines data model validation. Because business rules are the support-
ing documentation for a data model, it is practical for data administration
to present them to the business community when reviewing the data
model. Often business people do not understand a complex entity-relation-
ship diagram. They find following relationships between multiple associa-
tions entities difficult, but they can understand natural language. Unfortu-
nately, most automated tools do not support a presentation format that
integrates business rules with an entity-relationship diagram. Typically,
data administration members need to present the entity-relationship
diagram along with an accompanying report of business rules. One excep-
tion is ERWIN (from Logic Works), which can produce an entity-relation-
ship diagram that encompasses at least entity definitions.
BUSINESS RULE STORAGE AND AUTOMATION
Data administration may have greater success selling the use of business
rules when an automated mechanism for capturing them exists. Three basic
alternatives exist for automated business rule storage: word-processor,
computer-assisted software engineering (CASE) tool, and centralized
repository. A fourth alternative is a combination of some or all of these
basic alternatives.