Database Reference
In-Depth Information
when consolidating the partial data models into the global semantic data
model.
•
When the semantic data model is transformed into the appropriate con-
ventional data model—relational, hierarchical, or network—ensure that none
of the essential properties of the conventional data model is overlooked.
•
If your target database is relational, make sure that the transformed data model
conforms to all the relational rules and constraints.
DATA MODELING IN LOGICAL DESIGN
Think of logical design phase as consisting of two distinct steps—creation of the
semantic data model and transformation of the semantic data model into the desired
conventional model. Sometimes, semantic data modeling is taken to be part of the
requirements definition phase, perhaps considering the semantic model as the
output of the requirements definition phase. We would rather place semantic data
modeling in the logical database design phase. Both the semantic and the conven-
tional data models are conceptual or logical models. They are conceptual repre-
sentations of real-world information requirements.
Also, the skills needed for semantic data modeling and for gathering and defin-
ing information requirements are not the same. Analysts working on the require-
ments definition phase must possess special proficiency in analyzing business
operations, reviewing business processes, and interviewing users to gather the
requirements. Data modelers must have expertise in modeling techniques and
knowledge of leading CASE tools.
Steps for Completing Logical Design
In the previous section, we listed a number of critical factors to be concerned about
in the logical design process. Essentially, steps for completing the logical design
relate to the fulfillment of these factors. Semantic data modeling is the beginning
of the logical design phase. We have covered semantic data modeling in elaborate
detail in earlier chapters. You know the components of a semantic data model; you
are very familiar with data modeling techniques. Let us trace the steps to complete
the logical design from the semantic data model.
Figure 11-5 shows a semantic data model for a small florist business. Note the
components in the data model diagram. Observe the entities, attributes, and rela-
tionships. As this exercise is meant just to illustrate the steps for completing the logical
design, we are not considering all the possible entities and attributes for a typical
florist business. Let us begin with this data model and complete the logical design.
We will consider the relational data model as the conventional data model while
studying the steps for completing the logical design. Similar steps are applicable if
the conventional data models were to be hierarchical or network. Here are the
obvious basic steps to complete the logical design from the semantic data model:
•
Represent entities in the conventional data model.
•
Represent attributes in the conventional data model.
Search WWH ::
Custom Search