Normalization

Overview
What is normalization?

What are the normal forms?

How to normalize relations?

2
 Introduction
Question:
 How many, and what, relations (tables) should be
used to store my data?
 Is the current relation free of problems?
Stud_ID Stud_Name Course_ID Course_Name Instructor Office Room Credit
224 Waters CIS20 Intro CIS Greene CBA001 205G 5
224 Waters CIS40 Database Mgt Hong CBA908 311S 5
224 Waters CIS50 Sys.Analysis Purao CBA700 139S 5
351 Byron CIS30 COBOL Hong CBA908 629G 3
351 Byron CIS50 Sys.Analysis Purao CBA700 139S 5
421 Smith CIS20 Intro CIS Greene CBA001 205G 5
421 Smith CIS30 COBOL Hong CBA908 629G 3
421 Smith CIS50 Sys.Analysis Purao CBA700 139S 5

3
Normalization
Normalization is a process of producing a set
of related relations(tables) with desirable
attributes, given the data requirements of a
domain
 The goal is to remove redundancy and other data
modification (insertion, update and deletion)
problems
 Usually dividing a table into 2 or more tables
 Using Normal Forms as a guide

4
 Normal Forms
Normal forms are guidelines
(steps) for the normalization
process

DK/NF
...

5th Normal Form

ie s
om n

…
tio
all

bu
4th Normal Form (4NF)
r a liza

tm
a

Boyce Codd Normal Form (BC/NF)

bn
Fe orm

ore
3rd Normal Form (3NF)
N

tab
we

2nd Normal Form (2NF)

l es
1st Normal Form (1NF)
Unnormalized Form (UNF)
5
Normalization – 1NF
A table is in 1NF if
 it satisfies the definition of a relation
 Review: what are the characteristics of a
relation?
 No “repeating groups” (columns)

6
Repeating Groups
Customer First Telephone
Surname
ID Name Number
123 Robert Ingram 555-861-2025
555-403-1659
456 Jane Wright
555-776-4100
789 Maria Fernandez 555-808-9633

Customer First Tel. No.

Surname Tel. No. 1 Tel. No. 2
ID Name 3
123 Robert Ingram 555-861-2025
456 Jane Wright 555-403-1659 555-776-4100
789 Maria Fernandez 555-808-9633
Transforming to 1NF
Transforming to rows, rather than columns

Customer ID First Name Surname Telephone Number

123 Robert Ingram 555-861-2025
456 Jane Wright 555-403-1659
456 Jane Wright 555-776-4100
789 Maria Fernandez 555-808-9633

8
Transforming to 1NF: Example
Another example

UNF 1NF

9
Problems in 1NF
Basically has the same problem as the
spreadsheet tables
 Redundancy
 Data may be inconsistent after modification
Higher Normal Forms
Normal forms higher than 1NF deal with
functional dependency

Identifying the normal form level by

analyzing the functional dependency
between attributes (fields)

11
Functional Dependency
If each value of attribute A is associated with only
one value of attribute B, we say
 A determines B
 Or, B is dependent on A
 Denoted as: A  B

Functional dependence describes relationships

between attributes (not relations)

Note: A (or B) could be a set of fields

12
Functional Dependency Examples
Dependency example
 For each SSN, there is only one corresponding
first name (or last name), so:
 SSN determines FirstName
 SSN  FirstName

Non-dependency example
 Each instructor teaches multiple courses, so:
 InstructorId does not determines CourseNumber

13
Functional Dependency and Keys
By definition, a primary key (candidate key)
functionally determines all other attributes
 Primary key
 Surrogate key
 Composite primary key

Dependency diagram

ISBN Title PubDate ListPrice

14
Functional Dependency Exercise

CustomerNum  CustomerName?
{Street, City, State}  Zip?
CustomerName  Balance?
State (?) Zip
RepNum ( ? ) CustomerName

15
Normalization – 2NF
A relation is in 2NF, if
 It is in 1st normal formal, and
 All nonkey attributes must be functionally dependent on the
whole primary key (Full dependency)
 No partial dependency
 It implies that a relation is in 2NF if there is a single-
attribute primary key (candidate key)

Partial dependency
 A non-key attribute is dependent on part of a primary key

A B C D

16
A Relation in 1NF but Not in 2NF

Partial dependency
Transforming to 2NF
Identify primary key (PK)
 If PK consists of only one field, then it is in 2NF
 If PK is a composite key, then analyze functional dependency
between part of primary key and other non-key attributes
 Move partial dependency involved attribute to another relation

A B C D

A B C

B D
18
 Transforming to 2NF: Example
Order

Order_ID Order_Date Cust_ID Cust_Name Cust_Address

Order Value
1006 10/24/2004 2 Furniture Plano, TX
Furniture
1007 10/25/2004 6 Gallery Boulder, CO Ord_ID Prod_ID Ord_Quan
1006 7 2
Product Prod_ID Prod_Desc Prod_Finish Unit_Price OrderItem 1006 5 2
4 Entertainment Natural 650.00
Center Maple 1006 4 1
5 Writer's Desk Cherry 325.00 1007 11 4
7 Dining Table Natural Ash 800.00 1007 4 3
11 4-Dr Dresser Oak 500.00 19
Problems in 2NF
Again, redundancy and potential
inconsistency
Order_ID Order_Date Cust_ID Cust_Name Cust_Address
Value
1006 10/24/2004 2 Furniture Plano, TX
Furniture
1007 10/25/2004 6 Gallery Boulder, CO
Value
1008 11/1/2004 2 Furniture Plano, TX
Normalization – 3NF
A relation is in 3NF, if
 It is in 2nd normal formal, and
 All attributes must, and only, be functionally
dependent on candidate keys
 No transitive dependency

Transitive dependency
 A  B and B  C, then A  C

A B C D

21
A Relation in 2NF but Not in 3NF
Identify primary key (PK) and Look for
transitive dependence

Order_ID Order_Date Cust_ID Cust_Name Cust_Address

Value
1006 10/24/2004 2 Furniture Plano, TX
Furniture
1007 10/25/2004 6 Gallery Boulder, CO
Value
1008 11/1/2004 2 Furniture Plano, TX

Transitive dependency
22
 Transforming to 3NF
Move the attributes involved in transitive
dependency to another relation
Order_ID Order_Date Cust_ID Cust_Name Cust_Address
Value
1006 10/24/2004 2 Furniture Plano, TX
Furniture
1007 10/25/2004 6 Gallery Boulder, CO
Value
1008 11/1/2004 2 Furniture Plano, TX

Order Customer
Order_ID Order_Date Cust_ID Cust_ID Cust_Name Cust_Address
1006 10/24/2004 2 Value
1007 10/25/2004 6 2 Furniture Plano, TX
1008 11/1/2004 2 Furniture
6 Gallery Boulder, CO
23
Some Practical Tips
If there are attributes of two different entities in one table,
there are usually problems

To identify the normalization level, determine the primary key

first; then look for partial dependence and transitive
dependence

Generally relations in the 3rd normalization forms are considered

to be well formed; going higher may introduce unnecessary
structural complexity which is inefficient for data queries

Very often tables can go for lower normal forms (de-

normalization) depending on design requirements

24
Normalization Exercise 1-1

Which normal form is the above table in?

A. 1NF
B. 2NF
C. 3NF
D. UNF

25
Normalization Exercise 1-2

Which normal form is the above table (2nd one) in?

A. 1NF
B. 2NF
C. 3NF
D. UNF

26
Normalization Exercise 1-3

Which normal form is the above table (first one) in?

 1NF
 2NF
 3NF
 UNF

27
Normalization Exercise 1-4
Final database design with 3 tables

28
Summary
3NF: If the tables are in 2NF, and every non-key attribute is
dependent on the key, the whole key, and nothing but the key

Eliminate transitive
dependencies
2NF: If the tables are in 1NF, and every non-key attribute is
dependent on the key, the whole key

Eliminate partial
dependencies

1NF: If the tables are relations

Split repeating groups

in separate rows

UNF

29
Summary
Key concepts
 Normalization
 Normal forms
 Functional dependency