Database Reference
In-Depth Information
•
Decomposition of a relation into more than one relation or table is the general
technique in each step. Make sure that you do not lose any information when
decomposing a table into multiple tables.
•
Furthermore, at each step, you must be able to recreate the original set of rela-
tions from the set of decomposed relations. This will ensure that you did not
lose any information in the normalization step.
•
At the end of each step, it is a good practice to verify that the update, deletion,
and addition anomalies are eliminated in each of the resulting relations.
Normalization Example
In Chapter 7, we discussed the information requirements for the teaching aspects
of a university and developed an entity-relationship diagram as shown in Figure
7-22. The diagram represents the semantic data model. Later, in Chapter 9, we trans-
formed that semantic data model into a relational data model as shown in Figure
9-22. This was the data model transformation method for producing a true relational
data model.
Let us take the same example of information requirements for a university and
adopt the normalization approach to create a true relational model. First, create an
initial set of random tables intuitively from the information requirements. Figure
10-15 shows this initial data model.
Information Requirements
SocSecNo
StudntName
ClassNo
RoomNo
CourseDesc
Credits
FacltyName
Speczn
Department
Phone
TextbookTitle
Author
Possibility of multiple authors
Price
ExamName
Multiple exams for each student in a class
StudntScore
Score
MinScore
Major
Initial Set of Random Tables
SocSecNo
StudntName
ClassNo
RoomNo
CourseDesc
Credits
FacltyName
Speczn
Department
Phone
BookTitle
Author
Price
ExamName
Score
Grade
Major
Figure 10-15
University information: initial set of random tables.
Search WWH ::
Custom Search