Concepts of Database Management

Chapter 6
Database Design 2: Design Method

Objectives

• Discuss the general process and goals of database

design
• Define user views and explain their function
• Define Database Design Language (DBDL) and
use it to document database designs
• Create an entity-relationship (E-R) diagram to
visually represent a database design
• Present a method for database design at the
information level and view examples illustrating this
method

fb.me/SophornDS 1
 Objectives (continued)

• Explain the physical-level design process

• Discuss top-down and bottom-up approaches to
database design and examine the advantages and
disadvantages of both methods
• Use a survey form to obtain information from users
prior to beginning the database design process
• Review existing documents to obtain information
prior to beginning the database design

Objectives (continued)

• Discuss special issues related to implementing

one-to-one relationships and many-to-many
relationships involving more than two entities
• Discuss entity subtypes and their relationships to
nulls
• Learn how to avoid potential problems when
merging third normal form relations
• Examine the entity-relationship model for
representing and designing databases

fb.me/SophornDS 2
 Introduction

• Two-step process for database design

• Information-Level Design: completed
independently of any particular DBMS
• Physical-Level Design: information-level design
adapted for the specific DBMS that will be used
– Must consider characteristics of the particular DBMS

User Views

• User view: set of requirements necessary to

support operations of a particular database user
• Cumulative design: supports all user views
encountered during design process

fb.me/SophornDS 3
 Information-Level Design Method

• For each user view:

1. Represent the user view as a collection of tables
2. Normalize these tables
3. Identify all keys in these tables
4. Merge the result of Steps 1 through 3 into the
cumulative design

Represent the User View As a

Collection of Tables
• Step 1: Determine the entities involved and create
a separate table for each type of entity
• Step 2: Determine the primary key for each table
• Step 3: Determine the properties for each entity
• Step 4: Determine relationships between the
entities
– One-to-many
– Many-to-many
– One-to-one

fb.me/SophornDS 4
 • For example, suppose a requested user view
involves departments and employees, each
department can hire many employees, and
each employee can work in only one
department (a typical restriction).

Represent the User View As a

Collection of Tables (continued)
• One-to-Many relationship: include primary key of
the “one” table as a foreign key in the “many” table
• Many-to-Many relationship: create a new table
whose primary key is the combination of the
primary keys of the original tables
• One-to-One relationship: simplest implementation
is to treat it as a one-to-many relationship

10

fb.me/SophornDS 5
 Normalize the Tables

• Normalize each table

• Target is third normal form
– Careful planning in early phases of the process
usually rules out need to consider fourth normal form

11

Identify All Keys

• For each table, identify:

– Primary key
– Alternate keys
– Secondary keys
– Foreign keys
• Alternate key: column(s) that could have been
chosen as a primary key but was not
• Secondary keys: columns of interest strictly for
retrieval

12

fb.me/SophornDS 6
 Identify All Keys (continued)

• Foreign key: column(s) in one table that is

required to match value of the primary key for
some row in another table or is required to be null
– Used to create relationships between tables
– Used to enforce certain types of integrity constraints

13

Types of Primary Keys

• Natural key: consists of a column that uniquely

identifies an entity
– Also called a logical key or an intelligent key
• Artificial key: column created for an entity to serve
solely as the primary key and that is visible to users
• Surrogate key: system-generated; usually hidden
from users
– Also called a synthetic key

14

fb.me/SophornDS 7
 Database Design Language (DBDL)

• A mechanism for representing tables and keys

• Table name followed by columns in parentheses
– Primary key column(s) underlined
• AK identifies alternate keys
• SK identifies secondary keys
• FK identifies foreign keys
– Foreign keys followed by an arrow pointing to the
table identified by the foreign key

15

Database Design Language (DBDL)

(continued)

Employee (EmployeeNum, LastName, FirstName,

Street, City, State, Zip, WageRate,
SocSecNum, DepartmentNum)
AK SocSecNum
SK LastName, FirstName
FK DepartmentNum  Department

FIGURE 6-1: DBDL for the Employee table

16

fb.me/SophornDS 8
 Entity-Relationship (E-R) Diagrams

• Visually represents database structure

• Rectangle represents each entity
– Entity’s name appears above the rectangle
• Primary key for each entity appears above the line
in the entity’s rectangle
• Other columns of entity appear below the line in
rectangle

17

Entity-Relationship (E-R) Diagrams

(continued)
• Letters AK, SK, and FK appear in parentheses
following the alternate key, secondary key and
foreign key, respectively
• For each foreign key, a line leads from the
rectangle for the table being identified to the
rectangle for the table containing the foreign key

18

fb.me/SophornDS 9
 Entity-Relationship (E-R) Diagrams
(continued)

• Consider the following database design written in

DBDL:
Department (DepartmentNum, Name, Location)
Employee (EmployeeNum, LastName, FirstName,
Street, City, State, Zip, WageRate,
SocSecNum, DepartmentNum)
AK SocSecNum
SK LastName, FirstName
FK DepartmentNum  Department

19

Entity-Relationship (E-R) Diagrams

20

fb.me/SophornDS 10
 Merge the Result into the Design

• Combine tables that have the same primary key to

form a new table
• New table:
– Primary key is same as the primary key in the tables
combined
– Contains all the columns from the tables combined
– If duplicate columns, remove all but one copy of the
column
• Make sure new design is in third normal form

21

Merge the Result into the Design

FIGURE 6-3: Information-level design method

22

fb.me/SophornDS 11
 Database Design Examples

• Develop an information-level design

• Company stores information about sales reps,
customers, parts, and orders
• User view requirements
• Constraints

23

Database Design Examples (Cont.)

24

fb.me/SophornDS 12
 Database Design Examples (Cont.)

• User View 1 Requirements: For a sales rep, store

the sales rep’s number, name, address, total
commission, and commission rate. You will need to
create only one table to support this view:
Rep (RepNum, LastName, FirstName,
Street, City, State, Zip, Commission,
Rate)

25

Database Design Examples (Cont.)

• User View 2 Requirements: For a customer, store the
customer’s number, name, address, balance, and credit
limit. In addition, store the number and name of the
sales rep who represents this customer. A sales rep
can represent many customers, but a customer must
have exactly one sales rep. (A customer must have a
sales rep and cannot have more than one sales rep.)
Rep (RepNum, LastName, FirstName)
Customer (CustomerNum, CustomerName,
Street, City, State, Zip, Balance,
CreditLimit, RepNum)

26

fb.me/SophornDS 13
 Database Design Examples (Cont.)
• Or a single table:
Customer (CustomerNum, CustomerName,
Street, City, State, Zip, Balance,
CreditLimit, RepNum, LastName, FirstName)
• The table is in 2NF but it’s not in 3NF

27

Database Design Examples (Cont.)

28

fb.me/SophornDS 14
 Database Design Examples (Cont.)

• User View 3 Requirements: For a part, store the

part’s number, description, units on hand, item
class, number of the warehouse in which the part is
located, and price. All units of a particular part are
stored in the same warehouse.
Part (PartNum, Description, OnHand,
Class, Warehouse, Price)

29

Database Design Examples (Cont.)

30

fb.me/SophornDS 15
 Database Design Examples (Cont.)

• User View 4 Requirements: For an order, store

the order number; order date; number, name, and
address of the customer that placed the order; and
number of the sales rep who represents that
customer. In addition, for each line item within the
order, store the part number and description,
number of the part that was ordered, and quoted
price.

31

Database Design Examples (Cont.)

• The user also has supplied the following constraints:
a. Each order must be placed by a customer that is already
in the Customer table.
b. There is only one customer per order.
c. On a given order, there is, at most, one line item for a
given part. For example, part DR93 cannot appear in
several lines within the same order.
d. The quoted price might not match the current price in the
Part table, allowing the company to sell the same parts to
different customers at different prices. The user wants to
be able to change the price for a part without affecting
orders that are currently on file.

32

fb.me/SophornDS 16
 Database Design Examples (Cont.)

• Orders(OrderNum, OrderDate, CustomerNum,

CustomerName, Street, City, State, Zip,
RepNum, (PartNum, Description, NumOrdered,
QuotedPrice) )
• After Normalization process:
1. Orders(OrderNum, OrderDate, CustomerNum)
2. Customers(CustomerNum, CustomerName,
Street, City, State, Zip, RepNum)
3. Part(PartNum, Description)
4. OrderLine(OrderNum, PartNum, NumOrdered,
QuotedPrice)
33

Database Design Examples (Cont.)

34

fb.me/SophornDS 17
 Database Design Examples (Cont.)

35

36

fb.me/SophornDS 18
 Database Design Examples
(continued)
• Henry Books database: information about
branches, publishers, authors, and books
• User view requirements

FIGURE 6-9: DBDL for Book database after first user view

37

Database Design Examples

(continued)

FIGURE 6-10: DBDL for Book database after second user view

38

fb.me/SophornDS 19
 Database Design Examples
(continued)

FIGURE 6-13: Cumulative design after fifth user view

39

Physical-Level Design
• Undertaken after information-level design completion
• Most DBMSs support primary, candidate, secondary,
and foreign keys
• To enforce restrictions, DB programmers must
include logic in their programs

Employee (EmployeeNum,LastName,FirstName,
Street,City,State,Zip,WageRate,SocSecNum,
DepartmentNum)
AK SocSecNum
SK LastName
FK DepartmentNum  Department
40

fb.me/SophornDS 20
 Top-Down Versus Bottom-Up

• Bottom-up design method

– Design starts at low level
– Specific user requirements drive design process
• Top-down design method
– Begins with general database that models overall
enterprise
– Refines model until design supports all necessary
applications

41

Survey Form

• Used to collect information from users

• Must contain particular elements
– Entity information
– Attribute (column) information
– Relationships
– Functional dependencies
– Processing information

42

fb.me/SophornDS 21
 Obtaining Information from Existing
Documents
• Existing documents can furnish information
about database design
• Identify and list all columns and give them
appropriate names
• Identify functional dependencies
• Determine the tables and assign columns

43

Obtaining Information from Existing

Documents (continued)

44

fb.me/SophornDS 22
 Obtaining Information from Existing
Documents (continued)

List of possible attributes for the Holt Distributors invoice 45

Obtaining Information from Existing

Documents (continued)

46

fb.me/SophornDS 23
 Obtaining Information from Existing
Documents (continued)

47

Obtaining Information from Existing

Documents (continued)

FIGURE 6-19: Expanded list of entities

48

fb.me/SophornDS 24
 Obtaining Information from Existing
Documents (continued)
• Customer (CustomerNumber, CustomerSoldToName,
CustomerSoldToAddressLine1,
CustomerSoldToAddressLine2, CustomerSoldToCity,
CustomerSoldToState,
CustomerSoldToZip,CustomerRepNumber,
CustomerRepLastName, CustomerRepFirstName)
• Part (ItemNumber, ItemDescription, ItemPrice)
• Invoice (InvoiceNumber, InvoiceDate, OrderNumber,
ShipDate, Freight, InvoiceTotal)
• Order (OrderNumber, OrderDate, CustomerPONumber,
CustomerShipToName, CustomerShipToAddressLine1,
CustomerShipToAddressLine2, CustomerShipToCity,
CustomerShipToState, CustomerShipToZip)
• OrderLine (OrderNumber, ItemNumber,
ItemQuantityOrdered, ItemQuantityShipped,
ItemQuantityBackordered, ItemPrice) 49

One-to-One Relationship
Considerations
• Simply including the primary key of each table as a
foreign key in the other table
– No guarantee that the information will match
• One solution: create a single table
– Workable, but not the best solution
• Better solution
– Create separate tables for customers and sales reps
– Include the primary key of one of them as a foreign
key in the other

50

fb.me/SophornDS 25
 One-to-One Relationship
Considerations (continued)

FIGURE 6-20 One-to-one relationship implemented by including the primary key

of each table as a foreign key in the other

51

One-to-One Relationship
Considerations (continued)

FIGURE 6-21 Implementation of a one-to-one relationship in which information

does not match

52

fb.me/SophornDS 26
 One-to-One Relationship
Considerations (continued)

FIGURE 6-22 One-to-one relationship implemented in a single table

53

One-to-One Relationship
Considerations (continued)

FIGURE 6-23: One-to-one relationship implemented by including the primary

key of one table as the foreign key (and alternate key) in the other
table
54

fb.me/SophornDS 27
 Many-to-Many Relationship
Considerations
• Complex issues arise when more than two entities
are related in a many-to-many relationship
• Many-to-many-to-many relationship: involves
multiple entities
• Deciding between a single many-to-many-to-many
relationship and two (or three) many-to-many
relationships
– Crucial issue: independence

55

Many-to-Many Relationship
Considerations (continued)

56

fb.me/SophornDS 28
 Many-to-Many Relationship
Considerations (continued)

FIGURE 6-25: Result obtained by splitting the Sales table into three tables
57

Many-to-Many Relationship
Considerations (continued)

FIGURE 6-26: Result obtained by joining three tables—the second and third
rows are in error!
58

fb.me/SophornDS 29
 Nulls and Entity Subtypes

• Null
– Special value
– Represents absence of a value in a field
– Used when a value is unknown or inapplicable
• Splitting tables to avoid use of null values
• Entity subtype: table that is a subtype of another
table

59

Nulls and Entity Subtypes (continued)

FIGURE 6-27: Student table split to avoid use of null values

60

fb.me/SophornDS 30
 Nulls and Entity Subtypes (continued)

• Subtype called a category in IDEF1X terminology

• Incomplete category: records that do not fall into
the subtype
• Complete categories: all records fall into the
categories

61

Nulls and Entity Subtypes (continued)

FIGURE 6-29: Entity subtype in an E-R diagram

62

fb.me/SophornDS 31
 Nulls and Entity Subtypes (continued)

63

Nulls and Entity Subtypes (continued)

FIGURE 6-32: Two entity subtypes—incomplete categories

64

fb.me/SophornDS 32
 Nulls and Entity Subtypes (continued)

FIGURE 6-33: Two entity subtypes—complete categories

65

Avoiding Problems with Third Normal

Form When Merging Tables
• When combining third normal form tables, the
result might not be in third normal form
• Be cautious when representing user views
• Always attempt to determine whether determinants
exist and include them in tables

66

fb.me/SophornDS 33
 The Entity-Relationship Model

• An approach to representing data in a database

• Entities are drawn as rectangles
• Relationships are drawn as diamonds with lines
connecting the entities involved in relationships
• Composite entity: exists to implement a many-to-
many relationship
• Existence dependency: existence of one entity
depends on the existence of another related entity
• Weak entity: depends on another entity for its own
existence
67

The Entity-Relationship Model

(continued)

FIGURE 6-34: One-to-many relationship

68

fb.me/SophornDS 34
 The Entity-Relationship Model
(continued)

FIGURE 6-35: Many-to-many relationship

69

The Entity-Relationship Model

(continued)

FIGURE 6-36: Many-to-many-to-many relationship

70

fb.me/SophornDS 35
 The Entity-Relationship Model
(continued)

FIGURE 6-37: One-to-many relationship with attributes added

71

The Entity-Relationship Model

(continued)

FIGURE 6-38: Many-to-many relationship with attributes

72

fb.me/SophornDS 36
 The Entity-Relationship Model
(continued)

FIGURE 6-39: Composite entity

73

The Entity-Relationship Model

(continued)

FIGURE 6-40: Complete E-R diagram for the Premiere Products database
74

fb.me/SophornDS 37
 The Entity-Relationship Model
(continued)

FIGURE 6-41: E-R diagram with an existence dependency and a weak entity

75

The Entity-Relationship Model

(continued)
• Cardinality: number of items that must be included
in a relationship
– An entity in a relationship with minimum cardinality
of zero plays an optional role in the relationship
– An entity with a minimum cardinality of one plays a
mandatory role in the relationship

76

fb.me/SophornDS 38
 The Entity-Relationship Model
(continued)

FIGURE 6-43: E-R diagram that represents cardinality

77

Summary
• Database design is a two-part process:
information-level design (not dependent on a
particular DBMS), and physical-level design
(appropriate for the particular DBMS being used)
• User view: set of necessary requirements to
support a particular user’s operations
• Information-level design steps for each user view:
represent the user view as a collection of tables,
normalize these tables, represent all keys (primary,
alternate, secondary, and foreign), and merge the
results into the cumulative design
78

fb.me/SophornDS 39
 Summary (continued)

• Database design is represented in Database

Design Language (DBDL)
• Designs can be represented visually using entity-
relationship (E-R) diagrams
• Physical-level design process consists of creating a
table for each entity in the DBDL design
• Design method presented in this chapter is bottom-
up
• Survey form useful for documenting the information
gathered for database design process

79

Summary (continued)

• To obtain information from existing documents, list

all attributes present in the documents, identify
potential functional dependencies, make a tentative
list of tables, and use the functional dependencies
to refine the list
• To implement a one-to-one relationship, include
primary key of one table in the other table as a
foreign key and indicate the foreign key as an
alternate key

80

fb.me/SophornDS 40
 Summary (continued)
• If a table’s primary key consists of three (or more)
columns, determine whether there are independent
relationships between pairs of these columns
• If a table contains columns that can be null and the
nulls mean that the column is inapplicable for some
rows, you can split the table, placing the null
column(s) in separate tables
• The result of merging third normal form tables may
not be in third normal form
• Entity-relationship (E-R) model represents the
structure of a database using an E-R diagram
81

fb.me/SophornDS 41