SQIT3043

Data Modeling and Database

CHAPTER FIVE: DATA NORMALIZATION

5.1 THE PURPOSE OF NORMALIZATION

Normalization is a technique for producing a set of relations with desirable

properties, given the data requirements of an enterprise.
The characteristics of a suitable set of relations include:
- The minimal number of attributes necessary to support the data
requirements of the enterprise.
- Attributes with a close logical relationship (describe as functional
dependency) are found in the same relation.
- Minimal redundancy, with each attribute represented only once, with
the important exception of attributes that form all or part of foreign
keys, which are essential for the joining of related relations.
The benefits of using a database that has a suitable set of relations is that
the database will be easier for the user to access and maintain the data,
and take up minimal storage space on the computer.

5.2 HOW NORMALIZATION SUPPORTS DATABASE DESIGN

Normalization is a formal technique that can be used at any stage of

database design.
There are two approaches for using the normalization (Figure 5.1):
a. As a bottom-up standalone database design technique.
b. As a validation technique to check structure of relations
Figure 5.1 shows example of data sources that can be used for database
design. Although the users requirements specification (chapter 3) is the
preferred data source, it is possible to design a database based on the
information taken directly from other data sources such as forms and
reports. The same data sources can be used for both of the above
approaches.
However, in practice the approach taken is likely to be determined by the
size, extent and by the preference and expertise of the database
designer.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
1

SQIT3043

Data Modeling and Database

DATA SOURCE

Use top-down
approach such as ER
modeling

Users

Users
requirements
specification
Forms/reports
that are used or
generated by the
enterprise

APPROACH 2

Set of well-designed
relations

Use normalization
as a validation
technique to check
structure of
relations
(not covered)

Source
describing the
enterprise such
as data
dictionary &
corporate data
model
Use normalization
as a bottom-up
technique to create
set of relations.
(covered in this
chapter)

APPROACH 1
Figure 5.1 How normalization can be used to support database design

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
2

SQIT3043

Data Modeling and Database

5.3 DATA REDUNDANCY AND UPDATE ANOMALIES

The benefits of reducing data redundancy:

- Update to the data stored in the database are achieved with minimal
number of operations, thus reducing the opportunities for data
inconsistencies occurring in the database
- Reduction in the file storage space required by the base relation thus
minimizing costs.
Nevertheless, relational database also rely on the existence of a certain
amount of data redundancy in the form of copies of primary keys (or
candidate keys) acting as foreign keys in related relations to enable the
modelling of relationships between data.
Example of problems associated with unwanted data redundancy (Figure
5.2 and 5.3):
Staff
staffNo
SL21
SG37
SG14
SA9
SG5
SL41

sName
John White
Ann Beech
David Ford
Mary Howe
Susan Brand
Julie Lee

position
Manager
Assistant
Supervisor
Assistant
Manager
Assistant

salary
3000
12000
18000
9000
24000
9000

branchNo
B005
B003
B003
B007
B003
B005

Branch
branchNo

bAddress
22 Deer Rd, London
16 Argyll St, Aberdeen
163 Main St, Glasgow

B005
B007
B003

Figure 5.2 Staff and Branch relations

StaffBranch
staffNo

sName

position

salary

branchNo

bAddress

SL21
SG37
SG14
SA9
SG5
SL41

John White
Ann Beech
David Ford
Mary Howe
Susan Brand
Julie Lee

Manager
Assistant
Supervisor
Assistant
Manager
Assistant

3000
12000
18000
9000
24000
9000

B005
B003
B003
B007
B003
B005

22 Deer Rd, London

163 Main St, Glasgow
163 Main St, Glasgow
16 Argyll St, Aberdeen
163 Main St, Glasgow
22 Deer Rd, London

Figure 5.3 StaffBranch relation

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
3

SQIT3043

Data Modeling and Database

In Figure 5.3, the StaffBranch relation is an alternative format for the

Staff and Branch relations (figure 5.2). The relations have the
following form:
Staff
(staffNo, sName, position, salary, branchNo)
Branch
(branchNo, bAddress)
StaffBranch (staffNo, sName, position, salary, branchNo, bAddress)
Note: primary keys are underlined

In the StaffBranch relation, there is redundant data; the details of a

branch (i.e. branchNo and bAddress) are repeated for every member
of staff located at that branch.
In contrast, the branch details appear only once for each branch in
the Branch relation, and only branch number (branchNo) is repeated
in the Staff relation to represent where each member of staff is

located.
Relations that have redundant data may have problem called update
anomalies, which are classified as insertion, deletion or modification
anomalies.

5.3.1 Insertion anomalies

Based on figure 5.3, there are two main types of insertion anomalies:
- To insert the details of new members of staff into the
StaffBranch relation, we must include the details of branch
at which the staff are to be located. For example, to insert the
details of new staff located at branch B007, we must enter
the correct details of branch number B007 so that the
branch details are consistent with values for branch B007 in
other tuples of the StaffBranch relation. The relations shown
in figure 5.2 do not suffer from this potential inconsistency,
because we enter only the appropriate branch number for
each staff member in the Staff relation. Instead the details

of branch number B007 are recorded in the database as a

single tuple in the Branch relation.
To insert details of a new branch that currently has no
members of staff into the StaffBranch relation, it is necessary
to enter nulls into the attributes for staff such as staffNo.
However, as staffNo is the primary key for the StaffBranch
relation, attempting to enter nulls for staffNo violates entity

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
4

SQIT3043

Data Modeling and Database

integrity and is not allowed. We therefore cannot enter a

tuple for a new branch into the StaffBranch relation with a
null staffNo. The design of the relations shown in figure 5.2
avoids this problem, because branch details are entered in
the Branch relation separately from the staff details. The
details of staff ultimately located at that branch are entered
at a later date into the Staff relation.

5.3.2 Deletion anomalies

Based on figure 5.3, if we delete a tuple from the StaffBranch

relation that represents the last member located at a branch (e.g.
the tuples for staff number SA9 who is the only person working at
branch B007), the details about that branch are also lost from the
database.
The design of the relations in figure 5.2 avoids this problem because
Branch tuples are stored separately from Staff tuples and only the
attribute branchNo relates the two relations.

5.3.3 Modification anomalies

If we want to change the value of one of the attributes of a

particular branch in the StaffBranch relation, for example the
address for branch number B003, we must update the tuples of all
staff located at that branch.
If this modification is not carried out on all the appropriate tuples of
the StaffBranch relation, the database will become inconsistent. In
this case, the branch number B003 may appear to have different
addresses in different tuples.
The design of Staff and Branch relations in figure 5.2 have more
desirable properties that the StaffBranch relation in figure 5.3. It
shows that update anomalies can be avoided by decomposing the
original relation into the Staff and Branch relations.
There are two important properties associated with decomposition
of a larger relation into smaller relations:

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
5

SQIT3043

Data Modeling and Database

The lossless-join property: ensure that any instance of the

original relation can be identified from corresponding
instances in the smaller relations.
The dependency preservation property: ensures that a
constraint on the original relation can be maintained by
simply enforcing some constraint on each of the smaller
relations. In other words, we do not need to perform joins on
the smaller relations to check whether a constraint in the
original relation is violated.

5.4 FUNCTIONAL DEPENDENCIES

An important concept associated with normalization is functional

dependency which describes the relationship between attributes.

5.4.1 Characteristics of Functional Dependencies

Assume that a relational schema has attributes (A, B, C,Z) and that
the database is described by a single universal relation called R = (A,
B, C,Z). This assumption means that every attribute in the database
has a unique name.
Functional dependency describes the relationship between
attributes in a relation. For example, if A and B are attributes of relation
R, B is functionally dependent on A (denoted A B), if each value of
A is associated with exactly one value of B. (A and B may each consist
of one or more attributes).
If we know the value of A and we examine the relation that holds this
dependency, we find only one value of B in all the tuples that have a
given value of A, at any moment in time. Thus when two tuples have
the same value of A, they also have the same value of B. However, for
a given value of B, there may be several different value of A. This
dependency between attributes A and B can be represented
diagrammatically as shown in figure 4.4.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
6

SQIT3043

Data Modeling and Database

B is functionally dependent on A

Figure 5.4 A functional dependency diagram

When a functional dependency is present, the dependency is

specified as a constraint between attributes.
Hence we can describe the constraint between the two attribute as B
is functionally dependent on A or alternatively A functionally
determines B.
Attribute(s) on the left-hand side of the arrow (i.e. A) is called the
determinant.
Example 1 - Functional dependency:
- Consider the attributes staffNo and position in the Staff
relation in figure 5.2. For a specific staffNo, e.g. SL21, we can
determine the position of that member of staff as Manager. In
other words, staffNo functionally determine position as
shown in the figure 5.5 (a) below.
staffNo functionally determines position

staffNo

position

SL21

Manager

Figure 5.5 (a) staffNo functionally determines position (staffNoposition)

-

However, the opposite is not true. As shown in figure 5.5. (b),

position does not functionally determines staffNo. A member of
staff holds one position; however there may be several members
of staff with the same position.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
7

SQIT3043

Data Modeling and Database

position does not functionally determines staffNo

position

staffNo

Manager

SL21
SG5

Figure 5.5 (b) position does not functionally determines staffNo

-

The relationship between staffNo and position is one-to-one

(1:1), i.e. for each staff number there is only one position.
The relationship between position and staffNo is one-to-many
(1:*), i.e. there are several staff numbers associated with a given
position.
In this example, staffNo is the determinant of this functional
dependency.
For the purpose of normalization, we are interested in identifying
functional dependencies between attributes of relation that
have one-to-one relationship between the attribute(s) that
makes up the determinant on the left-hand side and the
attribute(s) on the right-hand side of a dependency.
Question: What is/are the functional dependency(ies) between
the attributes staffNo and sName?

An additional characteristics of functional dependencies that is useful

for normalization is that their determinant should have the minimal
number of attributes necessary to maintain the functional
dependency with the attribute(s) on the right-hand side. This
requirement is called full functional dependency.
Full functional dependency indicates that if A and B are attributes of a
relation, B is fully functionally dependent on A if B is functionally
dependent on A, BUT not on any proper subset of A.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
8

SQIT3043

Data Modeling and Database

Full functional dependency happens when removal of any attribute

from A results in the dependency no longer exist. If the dependency
still holds after removal of any attributes from A, the functional
dependency AB is called partially dependence.
Example 2 - Full functional dependency:
- Consider the following functional dependency that exist in the
Staff relation of figure 5.2:
staffNo, sName branchNo
- It is correct to say that each value of (staffNo, sName) is
associated with a single value of branchNo.
- However it is not a full functional dependency, because
branchNo is also functionally dependent on a subset of
(staffNo, sName), namely staffNo.In other words, the
functional dependency in the example is an example of a
partial dependency.
- The type of functional dependency that we are interested in
identifying is a full functional dependency as shown below:
staffNo branchNo
In summary, the functional dependencies that we use in normalization
have the following characteristics:
- There is one-to-one relationship between the attribute(s) on the
left hand-side (determinant) and those on the right-hand side of
a functional dependency. [Note: the relationship is in the
opposite direction that is from right-hand to the left-hand side
attributes can be one-to-one relationship OR one-to-many
relationship].
- They hold for all time.
- The determinant has the minimal number of attributes necessary
to maintain the dependency with the attribute(s) on the right
hand side. In other words, there must be a full functional
dependency between the attribute(s) on the left-hand and the
right-hand sides of the dependency.
There is additional type of functional dependency we called a
transitive dependency.
Transitive dependency: A condition where A, B, and C are attributes
of a relation such that if A B and B C, then C is transitively

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
9

SQIT3043

Data Modeling and Database

dependent on A via B (provided that A is not functionally dependent

on B or C).
Example 3 Transitive functional dependency:
- Consider the following functional dependencies within the
StaffBranch relation shown in figure 5.3:
staffNo sName, position, salary, branchNo,
bAddress
branchNo bAddress
- The transitive dependency branchNo bAddress exist on
staffNo via branchNo. In other words, the staffNo attribute
functionally determines the bAddress via the branchNo
attribute and neither branchNo nor bAddress functionally
determines stafNo.
- Transitive dependency is discussed in section 5.8.

5.4.2 Identifying Functional Dependency

Identifying all functional dependencies between a set of attributes

requires understanding about the meaning of each attribute and their
relationships this information may be obtained from the enterprise in
the form of discussion with users and/or appropriate documentation,
such as the users requirements specification.
However if the users cannot be consulted, the database designers
may need to use their common sense and/or experience to provide
the missing information.
Example 4 Identifying a set of functional dependencies for the
StaffBranch relation (Figure 5.3): (assume that the position held and
the branch determine a staff member salary)
- staffNo sName, position, salary, branchNo, bAddress
- branchNo bAddress
- bAddress branchNo
- branchNo, position salary
- bAddress, position salary
all the attributes on the right-hand side are functionally dependent
on the determinant on the left-hand side.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
10

SQIT3043

Data Modeling and Database

5.4.3 Identifying Primary Key for a Relation Using Functional Dependency

The main purpose of identifying functional dependencies for a relation

is to specify the set of integrity constraints that must hold on a relation.
An important integrity constraint to consider first is the identification of
candidate keys, one of which is selected to be the primary key for the
relation.
Example 5 Identifying the primary key for the StaffBranch relation.
- In example 4, we identify five functional dependencies for the
StaffBranch relation.
- The determinant for these five functional dependencies are:
staffNo, branchNo, bAddress, (branchNo, position)
and (bAddress, position)
- To identify the candidate keys for the StaffBranch relation, we
must identify the attribute (or group of attributes) that uniquely
identifies each tuple in this relation.
- Primary key is chosen from the candidate key(s) and all
attributes that are not part of the primary key (non-primary key
attibutes) should be functionally dependent on the key.
- For the StaffBranch relation, the only candidate key is

staffNo all other attributes of the relation functionally depend

on it.
Although there are other determinants in this relation, i.e.
branchNo, bAddress, (branchNo, position) and
(bAddress, position), they are not candidate keys for the
relation.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
11

SQIT3043

Data Modeling and Database

5.5 THE PROCESS OF NORMALIZATION

Normalization is a formal technique for analyzing relations based on their

primary key (or candidate keys) and functional dependencies (Codd,
1972b).
Involves a series of rules to test individual relations so a database can be
normalized to any degree.
When a requirement is not met, the relation violating the requirement must
be decomposed into relations that individually met the requirements of
normalization.
Three normal forms were initially proposed called First Normal Form (1NF),
Second Normal Form (2NF), Third Normal Form (3NF). Subsequently, R.
Boyce and E.F. Codd introduced a stronger definition of third normal form
called Boyce-Codd Normal Form (BNF) (Codd, 1974).
Higher normal forms (i.e. Fourth and Fifth Normal Form, 4NF and 5NF) were
introduce later which deals with situations that are very rare, hence are
not covered in this course.
For the relational data model, it is important to recognize that only First
Normal Form (1NF) that is critical in creating relations; all subsequent
normal forms are optional. However to avoid update anomalies, it is
generally recommended that we proceed to at least 3NF.
The normalization in this course is describe as a bottom up technique
extracting information about attributes from sample forms that are first
transformed into table format, which is described as being Unnormalized
form (UNF) a table that contains one or more repeating groups.
UNF is then subjected progressively to different requirements of each
normal form until ultimately the attributes shown in the original sample
forms are represented as a set of 3NF relations.
Although the example used in this chapter proceeds from a given normal
form to the one above, this is not necessarily the case with others. In some
cases a 1NF relation may be transformed directly to 3NF relations in one
step.
For each of the examples used in the coming sections, we assume that a
set of functional dependencies is given for each relation and that each
relation has a designated primary key (i.e. meaning of attributes and their
relationships is well understood before normalization begins).

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
12

SQIT3043

Data Modeling and Database

Data source

Users

Users
Requirement
Specification

Forms/reports
that are used
or generated
by the
enterprise

Transfer attributes into table format

Source
describing the
enterprise such
as data
dictionary and
corporate
model

(Described in section 5.6)

Unnormalized
Form (UNF)
Remove repeating groups

(Described in Section 5.6)

First
Normal Form
(1NF)
Remove partial dependencies

(Described in Section 5.7)

Second
Normal Form
(1NF)
Remove transitive dependencies

(Described in Section 5.8)

Third
Normal Form
(1NF)

Figure 5.6 Diagrammatic illustration of the process of normalization

MOHD. NOOR ABDUL HAMID
mohdnoor@uum.edu.my
13

SQIT3043

Data Modeling and Database

5.6 FIRST NORMAL FORM (1NF)

First Normal Form (1NF) A relation in which the intersection of each row
and column contains one and only one value.
We begin by transferring the data from the source (e.g. a standard data
entry form) into table format (Unnormalized table)
To transform unnormalized table to 1NF, we identify and remove repeating
groups within the table. There are two approaches to do so:
- By entering appropriate data in the empty columns of rows
containing the repeating data.
We fill in the blanks by duplicating the non-repeating data,
where required (i.e. flattening the table).
This approach introduces more redundancy into the original
UNF table as part of the flattening process.
- By placing the repeating data along with a copy of the original
key attribute(s), in a separate relation.
This approach is applied repeatedly in cases where there are
more than one repeating groups in an unnormalized table until
no repeating group remains.
This approach creates two or more relations with less
redundancy than in the original UNF table.
Example 6: First Normal Form
- Figure 5.7 shows a collection of DreamHome leases. For this
example we assume that a client rents a given property only once
and cannot rent more than one property at any one time.
DreamHome Lease
DreamHome Lease
DreamHome Lease
Client Number: CR76
(if known)
Full Name: John Kay
(Please Print)
Monthly Rent: 350

Property Number : PG4

Property Address: 6
Lawrence St, Glasgow
Owner Number: C040
(if known)

Rent Start: 01/07/07

Rent finish: 31/08/08

Full Name: Tina Murphy

(Please print)

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
14

SQIT3043

Data Modeling and Database

Figure 5.7 Collection of DreamHome leases.

Sample data is taken from the leases and is transformed into table
format (Unnormalized table) as shown in figure 5.8:

ClientRental
clientNo
CR76

cName
John
Kay

propertyNo
PG4

PG16
CR56

Aline
Stewart

PG4

PG36
PG16

pAddress
6
Lawrence
St,
Glasgow.
5 Novar
Dr.
Glasgow.
6
Lawrence
St,
Glasgow.
2 Manor
Rd,
Glasgow.
5 Novar
Dr,
Glasgow.

rentStart
1-Jul-07

rentFinish
31-Aug-08

rent
350

ownerNo
C040

oName
Tina
Murphy

1-Sep-08

1-Sep-09

450

C093

Tony
Shaw

1-Sep-06

10-June-07

350

C040

Tina
Murphy

10-Oct-07

1-Dec-08

375

C093

Tony
Shaw

1-Nov-09

10-Aug-10

450

C093

Tony
Shaw

Figure 5.8: ClientRental unnormalized table.

-

The key attribute for the ClientRental unnormalized table is

clientNo.
The repeating group in this table is:
Repeating Group = (propertyNo, rentStart, rentFinish,
rent, ownerNo, oName)
i.e. there are more than one values in the above fields for each
tuple.
To transform the unnormalize table into 1NF, we ensure there is only
a single value at each intersection of each row and column by
removing repeating group.

Using approach 1:
-

We remove repeating group by entering the appropriate client

data into each row as shown in figure 5.9:

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
15

SQIT3043

Data Modeling and Database

ClientRental
clientNo
CR76

cName
John
Kay

propertyNo
PG4

pAddress
6
Lawrence
St,
Glasgow.

rentStart
1-Jul-07

rentFinish
31-Aug-08

rent
350

ownerNo
C040

oName
Tina
Murphy

CR76

John
Kay

PG16

5 Novar
Dr.
Glasgow.

1-Sep-08

1-Sep-09

450

C093

Tony
Shaw

CR56

Aline
Stewart

PG4

6
Lawrence
St,
Glasgow.

1-Sep-06

10-June-07

350

C040

Tina
Murphy

CR56

Aline
Stewart

PG36

2 Manor
Rd,
Glasgow.

10-Oct-07

1-Dec-08

375

C093

Tony
Shaw

CR56

Aline
Stewart

PG16

5 Novar
Dr,
Glasgow.

1-Nov-09

10-Aug-10

450

C093

Tony
Shaw

Figure 5.9: 1NF ClientRental table

Using approach 2:
-

There are six functional dependencies for ClientRental relation:

a. FD1: (Primary Key)
clientNo, propertyNo rentStart, rentFinish
b. FD2: (Partial Dependency)
clientNo cName
c. FD3: (Partial Dependency)
propertyNo pAddress, rent, ownerNo, oName
d. FD4: (Transitive Dependency)
ownerNo oName
e. FD5: (Candidate Key)
clientNo, rentStart propertyNo, pAddress,
rentFinish,rent, ownerNo, oName.
f. FD6: (Candidate Key)
(propertyNo, rentstart) clientNo, cName,
rentFinish

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
16

SQIT3043

Data Modeling and Database

Composite keys for the relation are:

(clientNo, propertyNo), (clientNo, rentStart) and
(propertyNo, rentStar)
We select (clientNo, propertyNo)as the primary key for the
relationship.
The ClientRental relation is defined as follow:
ClientRental (clientNo, propertyNo, cName, pAddress,
rentStart, rentFinish, rent, ownerNo, oName)
The relation contains data describing clients, property rented, and
property owners which is repeated several times (i.e. data
redundancy).
Using second approach, we remove repeating group (e.g. client
names) by placing the repeating data along with a copy of
original key attribute (clientNo) in a separate relation as shown in
figure 5.10:

Client
clientNo
CR76
CR56

cName
John Kay
Aline Stewart

PropertyRentalOwner
clientNo
CR76

propertyNo
PG4

pAddress
6 Lawrence St,
Glasgow.

rentStart
1-Jul-07

rentFinish
31-Aug-08

rent
350

ownerNo
C040

oName
Tina
Murphy

CR76

PG16

5 Novar Dr.
Glasgow.

1-Sep-08

1-Sep-09

450

C093

Tony
Shaw

CR56

PG4

6 Lawrence St,
Glasgow.

1-Sep-06

10-June-07

350

C040

Tina
Murphy

CR56

PG36

2 Manor Rd,
Glasgow.

10-Oct-07

1-Dec-08

375

C093

Tony
Shaw

CR56

PG16

5 Novar Dr,
Glasgow.

1-Nov-09

10-Aug-10

450

C093

Tony
Shaw

Figure 5.10: Alternative 1NF Client and PropertyRentalOwner relations

-

The format of resulting 1NF relations are as follow:

Client (clientNo, cName)
PropertyOwnerRental (clientNo, propertyNo, pAddress,
rentStart, rentFinish, rent, ownerNo, oName)

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
17

SQIT3043

Data Modeling and Database

However this relation also contains some redundancy. To remove

this redundancy, the process need to be repeated.

5.7 SECOND NORMAL FORM (2NF)

2NF is based on the concept of full functional dependency where a

relation is in the 1NF and every non-primary key attribute is fully
functionally dependent on the primary key.
2NF applies to relations which composite keys (i.e. primary key composed
of two or more attributes). A relation with a single attribute primary key is
automatically in at least 2NF.
The normalization of 1NF relations to 2NF involves the removal of partial
dependencies by removing the partially dependent attributes from the
relation by placing them in a new relation along with a copy of their
determinant.
Example 7: Second normal Form
- There are six functional dependencies for ClientRental relation:
a. FD1: (Primary Key)
clientNo, propertyNo rentStart, rentFinish
b. FD2: (Partial Dependency)
clientNo cName
c. FD3: (Partial Dependency)
propertyNo pAddress, rent, ownerNo, oName
d. FD4: (Transitive Dependency)
ownerNo oName
e. FD5: (Candidate Key)
clientNo, rentStart propertyNo, pAddress,
rentFinish,rent, ownerNo, oName.
f. FD6: (Candidate Key)
(propertyNo, rentstart) clientNo, cName,
rentFinish

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
18

SQIT3043

Data Modeling and Database

Using these functional dependencies, we continue the process of

normalizing the ClientRental relation.
i. Test whether the relation is in 2NF:
- Identify any partial dependencies on the primary key:
cName is partially dependent on the primary key (i.e.
clientNo)
pAddress, rent, ownerNo, oName are partially dependent
on on the primary key (i.e. propertyNo).
rentStart and rentFinish are fully dependent on the
whole primary key.
Hence ClientRental relation is not in 2NF.
ii. If NOT, create new relations so that the non-primary key
attributes are removed along with a copy of the primary key on
which they are fully functionally dependent:
- This results in the creation of three relations in 2NF (Figure 5.11),
as every non-primary key attributes fully functionally dependent
on the primary key of the relation:
Client (clientNo, cName)
Rental (clientNo, propertyNo, rentStart,
rentFinish)
PropertyOwner (propertyNo, pAddress, rent,
ownerNo, oName)

Client
clientNo
CR76
CR56

cName
John Kay
Aline Stewart

Rental
clientNo
CR76
CR76
CR56
CR56
CR56

propertyNo
PG4
PG16
PG4
PG36
PG16

rentStart
1-Jul-07
1-Sep-08
1-Sep-06
10-Oct-07
1-Nov-09

rentFinish
31-Aug-08
1-Sep-09
10-June-07
1-Dec-08
10-Aug-10

PropertyOwner
propertyNo
PG4
PG16
PG36

pAddress
6 Lawrence St, Glasgow.
5 Novar Dr. Glasgow.
2 Manor Rd, Glasgow.

rent
350
450
375

ownerNo
C040
C093
C093

oName
Tina Murphy
Tony Shaw
Tony Shaw

Figure 5.11: 2NF relations derived from ClientRental relation

MOHD. NOOR ABDUL HAMID
mohdnoor@uum.edu.my
19

SQIT3043

Data Modeling and Database

5.8 THIRD NORMAL FORM (3NF)

3NF a relation is in the 1NF and 2NF and in which no non-primary key
attributes is transitively dependent on the primary key.
Involves the removal of transitive dependencies from the relation by
placing the attributes in a new relation along with a copy of the
determinant.
Example 8: Third normal Form
- The six functional dependencies for the Client, Rental and
PropertyOwner relations are:
a. Client
FD2: clientNo cName
b. Rental
FD1: clientNo, propertyNo rentStart, rentFinish
Primary key
FD5: clientNo, rentStart propertyNo, rentFinish
Candidate key
FD6: propertyNo, rentStart clientNo, rentFinish Candidate key
c. PropertyOwner
FD3: propertyNo pAddress, rent, ownerNo, oName
Primary key
FD4: ownerNo oName
Transitive dependency
-

The Client and Rental relations have no transitive dependencies,

hence are already in 3NF.
- oName transitively dependent on ownerNo. Hence PropertOwner
relation is not in 3NF. To transform it into 3NF, we remove this
transitive dependency by creating two new relations called
PropertyforRent and Owner (Figure 5.12).
- The resulting 3NF relations of the original ClientRental relation
have the form:
Client (clientNo, cName)
Rental (clientNo, propertyNo, rentStart,
rentFinish)
PropertyforRent (propertyNo, pAddress, rent, ownerNo)
Owner (ownerNo, oName)
The original ClientRental relation in figure 5.9 can be recreated by joining
the Client, Rental, PropertyforRent and Owner relations through
the primary key/foreign key mechanism.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
20

SQIT3043

Data Modeling and Database

Client
clientNo
CR76
CR56

cName
John Kay
Aline Stewart

Rental
clientNo
CR76
CR76
CR56
CR56
CR56

propertyNo
PG4
PG16
PG4
PG36
PG16

rentStart
1-Jul-07
1-Sep-08
1-Sep-06
10-Oct-07
1-Nov-09

rentFinish
31-Aug-08
1-Sep-09
10-June-07
1-Dec-08
10-Aug-10

PropertyForRent
propertyNo
PG4
PG16
PG36

pAddress
6 Lawrence St, Glasgow.
5 Novar Dr. Glasgow.
2 Manor Rd, Glasgow.

rent
350
450
375

ownerNo
C040
C093
C093

Owner
ownerNo
C040
C093
C093

oName
Tina Murphy
Tony Shaw
Tony Shaw

Figure 5.12: 3NF relations derived from ClientRental relation

5.9 GENERAL DEFINITIONS OF 2NF AND 3NF

2NF a relation that is in the first normal form and every non candidate
key attribute is fully functionally dependent on any candidate key.
3NF a relation that is in first and second normal form and in which no
non-candidate key attribute is transitively dependent on any candidate
key.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
21

SQIT3043

Data Modeling and Database

SOURCE:
Connoly, T., & Begg, C. (2010). Database Systems Apractical Approach to
Design, Implementation, and Management (5 ed.). Boston: Pearson.

REFERENCES:
Codd E.F. (1972b). Further Normalization of the Database Relational Model. In
Database Systems (Rustin R., ed.), Englewoods Cliffs, NJ: Prentice Hall.
Codd E.F. (1974). Recent Investigation in Relational Database Systems. In Proc.
IFIP Congress.

MOHD. NOOR ABDUL HAMID

mohdnoor@uum.edu.my
22