Building a Data Model - High-Performance Web Databases: Design, Development, and Deployment - page 222

• Employ a data-driven approach. Build a data model that represents

how the business uses and manages information, unbiased by any par-
ticular processing requirements or technological considerations.

• Incorporate both structural and integrity considerations in the data mod-

el.

Addressing structure alone might reveal how to organize records
within a database but will not ensure that the data values accurately
model business operations.

• Combine both conceptualization and normalization techniques into the

data modeling methodology. Conceptualization identifies entities, rela-

tionships, and identifiers (steps LDM1 through LDM6) and ensures
that the model on paper nub-represents the use of information within
the business. Normalization (step LDM7) ensures that the model is
structurally consistent and logical and has minimal syntactic redun-
dancy.

• Use diagrams to represent as much of the data model as possible. Dia-

grams are clearer, simpler, and more concise than text.

• Build a data dictionary to supplement the data model diagrams. Dia-

grams cannot convey everything. Textual specifications of entity/rela-
tionship/attribute definitions and business rules are also needed.

BENEFITS AND APPLICABILITY OF DATA MODELS

Data modeling is a critical prerequisite to effective database design. Log-
ical data modeling m also contribute to the success of other less obvious
endeavors, including:

• Assessing technological feasibility. A data-driven data model represents

data structures, relationships, and rules without any compromises to
accommodate technological limitations. The degree to which a partic-
ular technological implementation can materialize all components of
the data model provides an effective measure of the technology's op-
timality for related applications.

• Assessing software packages. In particular, the degree to which the da-

tabase design of an application software package approximates the
data model serves as one gauge of the package's ability to meet busi-
ness requirements.

• Assessing the impact of future changes. A data model provides a clear

picture of the underlying database design, unobliterated by techno-
logical detail. It therefore aids significantly in evaluating the effects of
changes in business requirements or technology.

• Facilitating end-user access to data. A data model is a business repre-

sentation of information structures and rules, and can be understood
by users. It can provide these users with an intelligible map of their
data within an end-user computing environment.