Chapter Three Entity-Relationship Data Modeling: Process and Examples

CHAPTER 3
ENTITY-RELATIONSHIP DATA MODELING: PROCESS AND
EXAMPLES

ANSWERS TO GROUP I QUESTIONS

3.1

List the steps in the data modeling process.

Plan Project
Determine Requirements
Specify Entities
Specify Relationships
Determine Identifiers
Specify Attributes
Specify Domains
Validate Model

3.2

Describe tasks to be accomplished during the planning step.

Major tasks in this step are obtaining project authorization and budget; building the project team;
planning the teams activities; establishing tools, techniques, and standards so that team members
will produce consistent results; and defining the projects scope. Scope definition is particularly
important for data modeling because everything is related to something else in most
organizations.

3.3

List common sources of requirements.

User Interviews
Observations of User Activity
Existing Forms and Reports
New Forms and Reports
Existing Manual Files
Existing Computer Files/Databases
Formally Defined Interfaces(XML)
Domain Expertise

3.4

Explain how user interviews and observations can be used to develop components of a
data model.
The goal of a user interview is to learn of potential entities and their relationships to one another.
It is also important to obtain source documents, such as existing forms and reports, which indicate
entities and their relationships and also provide lists of potential attributes. The goal in an
interview is to obtain an accurate description of that users or user groups perspective or external
view(s) of the data. Results from user interviews and observations will be documented as notes,
sketches, data element definitions, and so forth. Team members should create consistent
documentation and record their results into a consolidated repository.

3-1

Chapter Three Entity-Relationship Data Modeling: Process and Examples

3.5

Why is it important to clarify terms during user interviews?

Keep the possibility of terminology differences in mind when interviewing users.
CUSTOMER_HISTORY in accounting may be a very different thing from CUSTOMER_
HISTORY in sales. Thinking that you already know what CUSTOMER_HIS-TORY means
because you understand the accounting version of it is risky when you interview the sales
department.

3.6

How can forms and reports be used to develop components of the data model?
Existing documents can be analyzed to determine how data is currently viewed and processed,
and new documents indicate either changes in data views or the need for additional data to be
maintained.

3.7

How can existing files be used to develop components of the data model?
Existing files are excellent source of requirements. You can often determine entities and attributes
of entities by examining these files. For larger databases, data modeling products such as ERWin
and Visio can be used to reverse engineer an existing database to create a data model.

3.8

What does the term reverse engineer mean in the context of data models?
In the context of data modeling, the term reverse engineer means to that a current database in its
physical form or its internal schema and create a conceptual schema.

3.9

Explain how formal interfaces can be used to obtain requirements.

A formal interface is a standardized external view. Such interfaces are created by governmental
agencies, industry groups, and large enterprises to facilitate interorganizational data transfer.
Sometimes, new databases systems are required to receive data from or publish data to such an
interface. If so, the interface is a good source of data requirements.

3.10

Why is the use of formal interfaces increasing?

The need for interorganizational standards increased with the advent of the Internet. Additionally,
the use of XML makes the publication and use of such standards much easier. Almost every
business profession or major industry category has (or is in the process of developing) such
standards.

3.11

Describe a domain expert and explain advantages and disadvantages of using domain
experts in requirements specification.
Domain experts are individuals who have deep knowledge of a particular problem area.
For example, in a company such as 3M, there may be thousands upon thousands of abrasive
products. In this situation, there may be dozens of different ways to develop subtypes/supertype
relationships or category clusters. Domain experts can be invaluable for sorting through the
myriad of data modeling choices, eliminating many possibilities on the basis of their knowledge.
Using domain experts, however, has a downside. Nothing is more dangerous in a development
project than a group of experts sitting around a table saying, If I were a user, Id want No
one but a user knows what data users want or the form(s) in which they need it. So, although
domain expertise can shed light on a particular project, the primary focus for requirements
determination must be the users themselves.

3.12

In the following statement, what is probably an entity and what is probably an attribute?

3-2

Chapter Three Entity-Relationship Data Modeling: Process and Examples

I group the student applications by date

Student applications would probably be the entity while Date would probably the an attribute
3.13

In the following phrase, what is probably an entity and what is probably an attribute?
Can two entities be referenced here? If so, what are they?
I use the department name of professor
Professor is probably the entity while department name is probably the attribute. You would
probably not think of two entities being referenced here because of the wording department
name of professor. The of would imply that department name belong to (or is an attribute of)
professor. One might also imply that since the wording preceding the 'of' references a department
name, there may be a department entity.

3.14

Explain why the list of entities is likely to change as the data modeling project proceeds.
As the project progresses, this list will be modified many times. Some of the items in the list will
be found to be synonyms of one another; some items will be removed because they are found to
be irrelevant, and others will be added.

3.15

Why is it important to define the meaning and use of entities?

Sometimes different users use different names for the same entity. For example, when one refers
to a purchase, is this a purchase from the firm by a customer or is it a purchase from a supplier by
the firm? Defining the meaning and use of entities makes us determine if the are the same and if
they are not the same, create different names.

3.16

What elements comprise the definition of a relationship?

The definition of a relationship includes the identity of the parent and child entities, the
relationship type, the minimum and maximum cardinalities, and the name of the relationships.

3.17

Describe two ways to find relationships.

One is to examine every combination of two entities and determine whether a relationship might
exist. If so, next is to examine all requirements to determine the cardinalities and other
characteristics of the relationship. The second technique is to process all requirements documents
and find all relationships indicated in those documents. Sometimes, a combination of the two
approaches is used.

3.18

Explain in your own words why there is no such thing as a one-way relationship.
One thing cannot be related to a second thing with out the second thing being related to the first.
If I am you brother then (like it or not) you are my brother or sister. Even in a recursive
relationship where only one entity is involved, instances of the entity are related to each other.

3.19

Is it possible that only one direction of a relationship is used in an application? Why or

why not?
YES. Applications are designed for a specific function. For example, if there is a relationship
between student and class, an application would probably create a Class Roster by finding all
students in a class. That application would probably not be the same application that finds all
classes a student takes and creates a Student Schedule.

3-3

Chapter Three Entity-Relationship Data Modeling: Process and Examples

3.20

Give an example of an entity that has no obvious identifier. What should be done with
the entity?
A very simple example is a house or address. Initially you may think that the phone number of a
house would be an identifier but some houses may not have phones and all instances of the entity
must have an identifier. Next you may think house number and street name but what about rural
areas. These areas typically have Route Numbers but the houses do not have house numbers.
Then you could use Route Numbers and P.O. Box but houses in the city do not have P.O. Boxes.
(Note: The example above has played havoc with 911 service)
When this occurs, check to see if the entity is ID-dependent. If not, makes sure that it has all its
attributes and is not part of another entity.

3.21

Give an example, other than the one in this text, of two different entities that have the
same identifier.
Assume you are developing a database of a University. You find the entities Student, Professors,
Staff, and Parents. All four of these may have the same identifier, Social Security, Number.

3.22

Explain the relationship of an attribute and a domain.

Domains identify the characteristics of attributes. Each time an attribute is found it is assigned to
a domain. If it does not fit an existing domain a new domain is defined and the attribute is
assigned to the new domain.

3.23

How are domains created?

Domains are creates as lists of values allowed for an attribute or as a description of how the
attribute can be stored and any constraints placed on the attribute.

3.24

Explain the advantage of domain property inheritance.

When the domain properties change, all the attribute properties change as well. For example, if
the data type of a domain is changed, all attributes based on that domain will inherit the new data
type.

3.25

How can domains be used to enforce organizational data standards?

Some organizations develop a data dictionary of standardized domains and properties. These
domains and properties are imported into data modeling tools and are used to create attributes in
the data model. In this way, every data modeling project will be working with the same
standardized components.

3.26

Give an example of two attributes that suggest the need for a new entity. Choose one
whose name includes the word name and one whose name does not include name.
Attributes that are names of something other than the entity in which they reside can suggest the
need for an additional entity or relationship. For example, in a SALESPERSON entity, an
attribute such as Region suggests that there may be a need for a new entity named REGION. If
so, the Region attribute is replaced by the relationship between SALESPERSON and REGION.
Similarly, an entity PART with an attribute Category suggests the need for a CATEGORY entity
with a relationship to PART.

3.27

Why is it important to validate the data model?

3-4

Chapter Three Entity-Relationship Data Modeling: Process and Examples

This is perhaps the most important step because if the data model is incorrect, then the database
design will be incorrect.
3.28

Explain what is wrong with this statement: My data model is a better model of reality
than your data model.
We need to construct a data model that is the best model of the mental model that is in the minds
of the users of the system. Period. We need to know whether a model accurately reflects how the
users think about their world. The only relevant question when evaluating a data model is the
following: How well does it fit the mental models of the people who are going to use the
system? The person who is doing the data modeling may think the model under construction is a
weird way of viewing the world, but that is not the point. The only valid point is: Does it fit how
the users view their world?

3.29

If someone made the statement in question 3.28 to you, how would you respond?
The best model is the one that most closely fits fit how the users view their world.

3.30

Explain the following statement: We cannot say that human models are an accurate
representation of the real world because we know nothing about that world. We can only
say that those model work well enough that humans can think and communicate in a
fashion that has, so far, facilitated human species survival.
We know only what we have learned. We have learned only what we have been taught through a
variety of ways. Everything we do, including the way we model, is influenced by our personal
history. We have now way of knowing how our model would appear if our experiences were
changed.

3.31

What is the sole criterion for evaluating a data model?

How well does it fit the mental models of the people who are going to use the system?

3.32

Explain how a team review of a data model is conducted.

First, the data modeling group conducts design reviews among its own members. During
these reviews, the data model is evaluated against system requirements. During the reviews, the
team ensures that it will be possible to support user descriptions of the data needs. Additionally,
the team ensures that it will be possible to construct an external schema to support every form and
report in the requirements. During the team review, questions about the requirements are often
generated. After the data model is corrected to remove any problems discovered during the team
review, a second round of reviews is conducted with the users themselves.

3.33

Explain how a user review of a data model is conducted.

Usually one or two key users from each group that will use the new system are involved in these
reviews. Any questions that developed during the team review are answered during the user
review. Also, there are sometimes aspects of the data model that were decided by artistic
judgment rather than absolute requirement. There may, for example, be several different ways of
accomplishing the same goal. If so, sometimes the users are asked to comment upon the various
alternatives that could be used. If possible, the best way to conduct user interviews is to each the
users the meaning of the E-R symbols being used and then to ask users to review and comment
upon the data model in diagram format.

3-5

Chapter Three Entity-Relationship Data Modeling: Process and Examples

3.34

Explain how to use a mockup to demonstrate to users the consequences of maximum

cardinality between two entities.
An example order form that has space for only one salesperson manifests the fact that the
relationship from order to salesperson has a maximum cardinality of one. If the users ask, Where
do I put the second salesperson? it is likely that the maximum cardinality from order to
salesperson is greater than one.

3-6

Chapter Three Entity-Relationship Data Modeling: Process and Examples

ANSWERS TO GROUP II QUESTIONS

Answer the following questions using IDEF1X notation:
3.35

Examine the subscription form shown in Figure 3-28. Using the structure of this form, do
the following:
A. Create a single entity model. Specify the identifier and attributes.
SUBSCRIPTION
Address
Zip
Name
City
State
Subscription Length
Payment Status
First Issue
B. Create a model with two entities, one for customer and a second for subscription.
Specify identifiers, attributes, relationship name, type, and cardinalities.
CUSTOMER

SUBSCRIPTION

Name
Address
City
State
Zip

OrderBy /
SentTo

Subscription_Length
First_Issue
Payment_Status

C. Under what conditions do you prefer the model in A to that in B?

Model A would be the best model if a Customer were only allowed to have one subscription.
D. Under what conditions do you prefer the model in B to that in A?
Model B would be the best model if a Customer were only allowed to have one multiple
subscriptions.
3.36

Consider the traffic citation shown in Figure 3-29. The rounded corners on this form
provide graphical hints about the boundaries of entities represented here.
A. Create a data model with five entities. Use the data items on the form to specify
identifiers and attributes for those entities.

3-7

Chapter Three Entity-Relationship Data Modeling: Process and Examples

DRIVER
License_Id
Last_Name
First_Name
AddressStreet
City
State
Zip
Gender
BirthDate
Height
Weight
EyeColor
ViolationType
DriverSignature

TICKET
ViolationDay
ViolationMonth
ViolationYear
ViolationTime
Receives /
GivenTo

Dist
Detach
LocationMiles
LocationDirection
Location
Route/Street

VIOLATIONS
DescriptionOf /
ViolatedOn

ViolationDescription

Gives /
GivenBy
RegisteredBy /
RegisteredTo
OFFICER
PersonnelNumber

VEHICLE

Officer_Name

VehicleLicense
State
Color
Make
Type
VIN

B. Specify relationships among the entities. Name the relationship and give its type and
cardinalities. Indicate which cardinalities can be inferred from data on the form and
which need to be checked out with systems users.
Relationships, Types, and Cardinalities are show on the E-R Diagram. From the form you can
infer that a Driver register only one Vehicle. This is because of the single Registered Owner
in the Vehicle area. You can also imply that a Ticket can have several Violations because of
the multiple lines and the wording Violations (plural). You can imply that only one Officer
can give the Ticket because there is room for only one Officer Signature.
3.37

Examine the list of email messages in Figure 3-30. Using the structure and example
data items in this list, do the following:
A. Create a single entity data model for this list. Specify the identifier and all entities.
EMailMessage
MailID
From
Subject
Date
Size

Since no attribute is unique, a MailID was created.

3-8

Chapter Three Entity-Relationship Data Modeling: Process and Examples

B. Modify your answer to A to include entities SENDER and SUBJECT. Specify the
identifiers and attributes of entities and the type and cardinalities of relationships.
Explain which cardinalities can be inferred from Figure 3-30 and which need to be
checked out with users.
EMailMessage
SUBJECT

MailID
Date
Size

MessagesFor /
About

Subject

SENDER
SenderID

Sends /
SentBy

The one-to-many relationships between EMailMessage and Subject can be implied because
the subject repeats. For example, there are three messages with the Subject RE:Hotel.
The one-to-many relationships between EMailMessage and Sender can also be implied
because the sender repeats. For example, there are three messages from Tom Cooper.
C. There are two different styles of email address in the From column in Figure 3-30.
One style has the true email address; the second style (Tom Cooper, for example) is
the name of an entry in the users email dictionary. Create two categories of
SENDER for these two styles. Specify identifiers and attributes.
SUBJECT

EMAILMESSAGE
Mail_Id
Date
Size

Subject

For /
About
Sends /
SentBy

SENDER
TrueAddress

DICTIONARYENTRY

EMAILADDRESS

LogicalName

Logical_Name

3-9

Chapter Three Entity-Relationship Data Modeling: Process and Examples

3.38

Examine the list of stock quotations in Figure 3-31. Using the structure and example
data items in this list, do the following:
A. Create a single entity data model for this list. Specify the identifier and attributes.
STOCKQUOTE
Symbol
Name
LastQuote
Change
PercOfChange

B. Modify your answer to A to include the entities COMPANY and INDEX. Specify the
identifier and attributes of the entities and the type and cardinalities for relationships.
Explain which cardinalities can be inferred from Figure 3-31 and which need to be
checked out with users.
COMPANY

INDEX
Symbol

Symbol

Index_Name

Name

Lists /
ListedIn

Valued /
ValurFor
1

STOCKQUOTE
Change
PercOfChange

Each of these relationships is a One-to-1. A Quote is for a single Index or Company and an Index
or Company has one quote. They are Identifying relationships because the key to all three tables
is the same, the Symbol. This is a unique situation because a Stockquote will belong to either an
Index or a Company but not both. All of these can be implied by the data shown.
C. The list in Figure 3-31 is for a quotation on a particular day at a particular time of
day. Suppose that the list were changed to show closing daily prices for each of
these stocks and that it includes a new column: QuoteDate. Modify your model in B
appropriately.

3-10

Chapter Three Entity-Relationship Data Modeling: Process and Examples

COMPANY

INDEX
Symbol

Symbol

Index_Name

Name

Lists /
ListedIn

Valued /
ValurFor
P

STOCKQUOTE
QuoteDate
Change
PercOfChange

The only things that would change are the identifier and the maximum cardinality for the
weak entity StockQuote. The identifier would now need to include the QuoteDate and the
cardinality would be 1:N rather than 1:1.
D. Change your model in C to include the tracking of a portfolio. Assume the portfolio
has an owner name, phone, email address, and a list of stocks held. The list
includes the identity of the stock and the number of shares held. Specify all
additional entities, their identifiers and attributes, and the type and cardinality of all
relationships.
PORTFOLIO
OwnerName
Phone
Email

STOCKPURCHASE
Buys /
SoldTo

NumberOfShares
SoldTo /
BoughtFrom

INDEX

SoldTo /
BoughtFrom
COMPANY

Symbol

Symbol

Index_Name

Name

Lists /
ListedIn

Valued /
ValurFor
P

STOCKQUOTE
QuoteDate
Change
PercOfChange

E. Change your answer to question D to keep track of portfolio stock purchases and
sales in a portfolio. Specify entities, their identifiers and attributes, and the type and
cardinality of all relationships.

3-11

Chapter Three Entity-Relationship Data Modeling: Process and Examples

PORTFOLIO
OwnerName
Phone
Email

STOCKPURCHASE
Buys /
SoldTo

SoldTo /
BoughtFrom SoldTo /
BoughtFrom

Sells /
SoldBy
INDEX

STOCKSALE
NumberOfShares

NumberOfShares

BoughtFrom /
SoldTo

COMPANY

Symbol

Symbol

Index_Name
SoldTo /
BoughtFrom
Lists /
ListedIn

Name
Valued /
ValurFor
P

STOCKQUOTE
QuoteDate
Change
PercOfChange

3.39

Figure 3-32 shows the specifications for single stage air compressor products. Notice
that there are two categories of products for different Air Performance: the A models are
at 125 (pounds per square inch of pressure), and the E models are at 150 (pounds per
square inch of pressure). Using the structure and example data items in this list, do the
following:
A. Create a category cluster to represent these compressors. The generic entity will
have attributes for all single stage compressors, and the category entities will have
attributes for products having the two different types of Air Performance. Assume
there might be additional products with different types of Air Performance. Specify
the entities, identifiers, attributes, relationships, type of category cluster, and possible
determinant.

3-12

Chapter Three Entity-Relationship Data Modeling: Process and Examples

AIRCOMPRESSOR
Model
HP
Tank
AppShipWt
Length
Weight
Height

125LB
PumpRPM
CFMDisp
DELDAir

150LB
PumpRPM
CFMDisp
DELDAir

The Entities, Identifiers, Attributes, relationships are sown in the diagram above. This is a
complete category cluster. There does not appear to be a determinate.
B. Figure 3-33 shows a different model for the compressor data. Explain the entities,
their type, the relationship, its type, and its cardinality. How well do you think this
model fits the data shown in Figure 3-32?
The SS_COMPRESSOR entity is a strong entity and the AIR_PERFORMANCE_TYPE
entity is a weak entity. The relationship is one-to-many since a compressor may have more
than one type. The relationship is an identifying connection relationship since the identifier
for a type requires the Model and the AirPerformance to be unique. Each type must belong to
an air compressor but an air compressor does not have to have a type. Both models (A. and
B.) fit the data show.
C. Compare your answer in question A to the model in Figure 3-33. What are the
essential differences in the two models? Which do you think is better?
The essential difference is that A. models the Type as a category cluster with a maximum of
two types. The model will allow additional types because it is incomplete but the additional
types cannot have the PumpRPM, CFMDisp, and DELDAir attributes. Model B. models the
Type as a weak entity with an unlimited number of types. I would prefer the model in B.
because it will allow you to have more than two Types.
D. Suppose you had the job of explaining the differences in these two models to a
highly motivated, intelligent end user. How would you accomplish this?
I would focus the discussion on the ability of model B. to allow an unlimited number of air
compressor types. This makes the model much more versatile.
3.40

Figure 3-34 shows a listing of movie times at theaters in Seattle. Using this data as an
example, do the following:
A. Create a model to represent this report using the entities MOVIE, THEATER, and
SHOW_TIME. Assume that theaters may be showing movies other than this one.
Although this report is for a particular day, your data model should allow for movie

3-13

Chapter Three Entity-Relationship Data Modeling: Process and Examples

times on different days as well. Specify the identifier of the entities and their
attributes. Name the relationships, describe the relationships type, and indicate
relationship cardinality.
THEATER

SHOWTIMES

MOVIE
Movie_Name
MovieDescription

WhenShown /
WhatShown

Day
Time

Theater_Name
Shown /
ShownAt

Distance
Street
City
Phone

Relationship Types are identifying and cardinalities are 1-to-Many.

Explain which cardinalities you can logically deduce from Figure 3-34 and which
need to be checked out with users. Assume that distance is an attribute of
THEATER.
From the report you can tell that a Move can be show at many different Theaters. You cannot
tell that a theater may show several different movies. This cardinality would need to be
checked out with users. You can also tell that Movies are shown at different times.
B. This report was prepared for a user who is located near downtown Seattle. Suppose
that it is necessary to produce this same report for these theaters, but for a user
located in a suburb such as Bellevue, Renton, Redmond, and Tacoma (these are
suburbs of Seattle). In this case, distance cannot be an attribute of THEATER.
Change your answer in A for this situation. Specify the identifier of entities and their
attributes. Name the relationships, describe the relationships type, and indicate
relationship cardinality.
SUBURB
Area_Name

DISTANCE
DistanceTo /
DistanceFrom

MovieDescription

THEATER

SHOWTIMES

MOVIE
Movie_Name

Distance

WhenShown /
WhatShown

Day
Time

DistanceTo /
DistanceFrom

Theater_Name
Shown /
ShownAt

Street
City
Phone

The new relationships are Identifying relationships.

C. Now, suppose that you want to make this data model national. Change your answer
to question B so that it can be used for other metropolitan areas. Change your

3-14

Chapter Three Entity-Relationship Data Modeling: Process and Examples

answer in A for this situation. Specify the identifier of entities and their attributes.
Name the relationships, describe the relationships type, and indicate relationship
cardinality.
A. and B. Both models would be done as shown below.
SUBURB
Area_Name
State_Name

DISTANCE
DistanceTo /
DistanceFrom

MovieDescription

Distance

THEATER

SHOWTIMES

MOVIE
Movie_Name

DistanceTo /
DistanceFrom

WhenShown /
WhatShown

Day
Time

Theater_Name
Shown /
ShownAt

Street
City
Phone

The only change that would need to be made is to add the State to the key of the Suburb. This is
done in case the same suburb name is used to more than one state.
D. Modify your answer to question C to include the leading cast members. Assume that
the role of a cast member is not to be modeled. Specify the identifier of new entities
and their attributes. Name new relationships, describe their relationships type, and
indicate their cardinality.

3-15

Chapter Three Entity-Relationship Data Modeling: Process and Examples

SUBURB
Area_Name
State_Name

DISTANCE
DistanceTo /
DistanceFrom

Distance

DistanceTo /
DistanceFrom

LEADINGCASTS
Cast_Name

PlaysIn /
Stars

Movie_Name
MovieDescription

THEATER

SHOWTIMES

MOVIE
WhenShown /
WhatShown

Day
Time

Theater_Name
Shown /
ShownAt

Street
City
Phone

Since a cast member can be in many movies and a move can have many leading castmembers, the
relationship is Many-to-Many or a Non-specific connection relationship.
3.41

Consider the three reports in Figure 3-35. The data are samples of data that would
appear in the reports like these.
A. Make a list of as many potential entities as these reports suggest to you.
Issuer
Ingredient
Supplier
Product
B. Examine your list to determine whether any entities are synonyms. If so, consolidate
your list.
It is possible that Issuer and Supplier are synonyms. This cannot be determined from the
report data so would need to check with the user.
C. Construct an IDEF1X diagram showing relationships among your entities. Name
each relationship and specify cardinalities. Indicate which cardinalities you can justify
on the basis of these reports and which you will need to check out with the users.

3-16

Chapter Three Entity-Relationship Data Modeling: Process and Examples

PRODUCT

ISSUER
IssuerName

Produces /
ProducedBy

ProductName

INGREDIENT
Uses /
UsedIn

IngredientName

Costs /
SellsFor
PriceList

SUPPLIER
SupplierName

3.42

Sells /
Charges

Price

Consider the CD cover in Figure 3-36.

A. Specify identifiers and attributes for the entities CD, ARTIST, ROLE, and SONG.
CD (Title, Comment, BookBy, MusicBy, LyricsBy, ProducedBy, CoreographedBy)
ARTIST (ArtistName)
SONG (Title, Time)
ROLE (Role)
B. Construct an IDEF1X diagram showing relationships among these three entities.
Name each relationship and specify cardinalities. Indicate which cardinalities you
can justify on the basis of the CD cover and which you will need to check out with
the users.
CD
CDTitle
Comment
BookBy
MusicBy
LyricsBy
DirectedBy
CoreographedBy

SONG
Title
Time

ARTIST
ArtistName

ROLE
Role

The cardinalities seem to be apparent. The only one I could question is the cardinality
between CD and Song. Could a Song appear on more than one CD.
C. Consider a CD that does not involve a musical, so there is no need for ROLE. There
is, however, a need for an entity SONG_WRITER. Create an IDEF1X diagram for
CD, ARTIST, SONG, and SONG_WRITER. Assume that an ARTIST can either be a
group or an individual. Assume that some artists record individually and as part of a
group.

3-17

Chapter Three Entity-Relationship Data Modeling: Process and Examples

ARTIST
ArtistName

CD
CDTitle
Comment
BookBy
MusicBy
LyricsBy
DirectedBy
CoreographedBy

SONG

ARTISTGROUP

Title

GroupName

Time

1
SONGWRITER
WriterName

D. Combine the models you developed in your answers to questions B and C. Create
new entities if necessary, but strive to keep your model as simple as possible.
Specify identifiers and attributes of new entities, name new relationships, and
indicate their cardinalities.
ROLE

ARTIST

Role

ArtistName
1

CD
CDTitle
Comment
BookBy
MusicBy
LyricsBy
DirectedBy
CoreographedBy

SONG

ARTISTGROUP

Title

GroupName

Time

1
SONGWRITER
WriterName

ANSWERS TO MODELING PROBLEMS

3.43 The Jefferson Dance Club teaches social dancing and offers both private and group
lessons. Jefferson charges $45 per hour per student (or couple) for a private lesson and
charges $6 per hour per student for a group lesson. Private lessons are offered six days
per week (from noon until 10 p.m.). Group lessons are offered in the evenings.

3-18

Chapter Three Entity-Relationship Data Modeling: Process and Examples

Jefferson employs two types of instructors: full-time salaried instructors and part-time
instructors. The full-time instructors are paid a fixed amount per week, and the part-time
instructors are paid either a set amount for an evening or a set amount for teaching a
particular class.
In addition to the lessons, Jefferson sponsors two weekly social dances featuring
recorded music. The admission charge is $5 per person. The Friday night dance is the
more popular and averages around 80 people; the Sunday night dance attracts about 30
attendees. The purpose of the dances is to give the students a place in which to practice
their skills. No food or drinks are served.
Jefferson wants to develop an information system to keep track of students and the
classes they have taken. Jeffersons managers also want to know how many and which
types of lessons each teacher has taught and to compute the average revenue generated
per lesson for each of their instructors.
Suppose Jefferson has hired you to construct a database for them. To do this, you decide
to create a data model. You know you will need to interview users and gather forms,
reports, and other requirements. Before you do this, however, you decide to construct a
trial data model. You are hoping that this model will help you determine specific questions
that you will need to ask Jefferson personnel.
A. To construct such a trial model, read through the preceding description and pick out all
of the important nouns. From these nouns, construct a list of potential entities.
Nouns
club
hour
evening
fixed amount
admission
Sunday
drinks
them
forms
personnel

lessons
couple
instructor
set amount
person
attendees
manager
you
reports

private lesson
student
full-time-salaried instructor
dance
Friday
skill
revenue
data model
requirements

group lesson
week
part-time instructor
music
people
food
database
user
question

Entities
lessons
instructor
person
user

private lesson
dance
attendee

group lesson
full-time-salaried instructor
skill

student
part-time instructor
manager

B. Review the list you made in A and eliminate any synonyms. Also, identify entities that
are subtypes or category entities. Name a likely identifier for each entity. Eliminate any
hard-to-justify entities.
Lesson: Id = LessonID
Group lesson: Subtype of Lesson
Instructor: Id = Instructor Name
part-time instructor: Subtype of Instructor

3-19

Private lesson: Subtype of Lesson

Student: Id = Student Name
full-time-salaried instructor: Subtype of Instructor
Attendee: Attendee Name

Chapter Three Entity-Relationship Data Modeling: Process and Examples

C. This description does not provide sufficient data for you to specify attributes of the
entities you have identified. Therefore, create an IDEF1X diagram that shows only
entities and relationships. Name each relationship and specify its cardinalities. Justify
the cardinality decisions you make.
ATTENDEE

Attends /
AttendedBy

DANCE

Attends /
AttendedBy

STUDENT
Takes /
IsTakenBy

Teaches /
TaughtBy

INSTRUCTORS

PARTTIME

FULLTIME

LESSON

PRIVATE

GROUP

D. Now, examine your model and make a list of questions that you would need to ask in
order to create an accurate data model.
1. Can people other than students attend a dance. If no, then ATTENDEE can be eliminated.
2. Can more than on instructor teach a lesson. This affects the cardinality of the IsTaughtBy
relationship.
3. Are there any kind of Instructors other than part-time and full-time.
E. Do you think the process of creating a trial data model before interviewing users is a
good idea? What are the advantages? What are the disadvantages?
It is definitely a good idea. You can get answers more quickly this way. It also gets the users
involved in the process much earlier. The only disadvantage is the extra time required from users
but the advantages far outweigh this disadvantage.
3.44San Juan Sailboat Charters is an agent that leases sailboats to customers for a fee. San
Juan does not own any sailboats; it leases boats on behalf of the owners who wish to
earn income when they are not using their boats. San Juan charges a fee for its service.
San Juan specializes in boats that can be used for multi-day or weekly chartersthe
smallest sailboat in its inventory is 28 feet and the largest is 51 feet.
Each sailboat is fully equipped at the time it is leased. Most of the equipment is provided
by the owners, but some is added by San Juan. The owner-provided equipment includes
what is fixed on the boat, such as radios, compasses, depth indicators and other
instrumentation, stoves, and refrigerators. Other owner-provided equipment is not
installed as part of the boat. Such equipment includes sails; lines; anchors; dinghies; life
preservers; and (in the cabin) dishes, silverware, cooking utensils, bedding, and the like.
San Juan provides consumable equipment, (which can also be considered supplies)
such as charts, navigation books, tide and current tables, soap, dishtowels, toilet paper,
and similar items.

3-20

Chapter Three Entity-Relationship Data Modeling: Process and Examples

An important part of San Juans responsibilities is keeping track of the equipment on the
boat. Much of it is expensive, and some of it, particularly what is not attached to the
boat, can easily be lost or stolen. Customers are responsible for all equipment during
the period of their charter. San Juan likes to keep accurate records of its customers and
the charters, not only for marketing but also for recording the trips that customers have
taken. Some itineraries and weather conditions are more dangerous than others, so San
Juan likes to know which customers have what experience.
Most of San Juans business is bare-boat chartering, which means that no skip-per or
other crew is provided. In some cases, however, customers request the services of a
skipper or other crew member, so San Juan hires such personnel on a part-time basis.
Sailboats often need maintenance. San Juan is required by its contracts with the boat
owners to keep accurate records of all maintenance activities and costs, including
normal activities (for example, cleaning or engine-oil changes) and unscheduled repairs.
In some cases, repairs are necessary during a charter.
A boat engine, for example, might fail while the boat is far away from San Juans facility.
In this case, the customers radio the San Juan dispatcher, who determines the best
facility to make the repair and sends the facilitys personnel to the disabled boat. To
make these decisions, the dispatchers need information about repair facilities as well as
past histories of repair quality and costs.
Suppose that San Juan has hired you to construct a database for them. To do this, you
decide to create a data model. You know you will need to interview users and gather
forms, reports, and other requirements. Before you do this, however, you decide to
construct a trial data model. You are hoping that this model will help you determine
specific questions that you will need to ask Jefferson personnel.
A.

B.

To construct such a trial model, read through the preceding description and pick out
all of the important nouns. From these nouns, construct a list of potential entities.
Nouns
agent
owners
feet
trip
crew
unscheduled repair
personnel

sailboat
income
equipment
itineraries
personnel
repair
history

customer
boat
owner-provided equipment
weather condition
contract
dispatcher

fee
charter
consumable equipment
skipper
maintenance activity
facility

Entities
Sailboat
Equipment
Personnel
Repair

customer
skipper
contract
facility

owner
owner-provided equipment
maintenance activity
personnel

charter
consumable equipment
unscheduled repair

Review the list you made in A and eliminate any synonyms. Also, identify entities that
are subtypes or category entities. Name a likely identifier for each entity. Eliminate any
hard-to-justify entities.
Sailboat: Id = BoatID

Customer: Id = Customer_Id

3-21

Chapter Three Entity-Relationship Data Modeling: Process and Examples

Owner: Id = Owner_Id
Charter: Id = BoatId and Date
Equipment: Id = BoatId and Description
Contract: Id = BoatId
Repair: Id = BoatId and RepairId
Facility: Id = FacilityName
maintenance activity: Subtype of repair
unscheduled repair: Subtype of repair
Skipper: ID = SkipperID
Personnel Id = CrewId
owner-provided equipment: Subtype of equipment
consumable equipment: Subtype of equipment
C.

This description does not provide sufficient data for you to specify attributes of the
entities you have identified. Therefore, create an IDEF1X diagram that shows only
entities and relationships. Name each relationship and specify its cardinalities. Justify
the cardinality decisions you make.
OWNERPROVIDED

CONSUMABLE

SKIPPER

OWNER
EQUIPMENT
AgreesTo /
SignedBy
CONTRACT

IsOn /
SailedBy

IsOn /
Carries
AssignedTo /
IsFor

SAILBOAT

Sails /
SailedBy

AssignedTo /
IsFor

CHARTER

Signs /
SignedBy

PERSONNEL

CUSTOMER

Requires /
MakeTo
FACILITY

Makes /
MadeBy

REPAIR

UNSCHEDULED

D.

MAINTENANCE

Now, examine your model and make a list of questions that you would need to ask in
order to create an accurate data model.
1. Does Sam Juan make repairs or are all repairs made by other Facilities. This affects the
minimum cardinality between Facility and Repair.
2. Are Skippers a set of people different than other crew personnel. If a person may be a
skipper on a crew for one charter and simply be one of the crew on a different charter,
Skippers and Personnel will be combined into one entity. That entity would have two
relationships with charter.

E.

Do you think the process of creating a trial data model before interviewing users is a
good idea? What are the advantages? What are the disadvantages?

3-22

Chapter Three Entity-Relationship Data Modeling: Process and Examples

It is definitely a good idea. You can get answers more quickly this way. It also gets the users
involved in the process much earlier. The only disadvantage is the extra time required from users
but the advantages far outweigh this disadvantage.

ANSWERS TO FIREDUP PROJECT QUESTIONS

Review the FiredUp Project Questions at the end of Chapter 2 and answer the questions there
if you have not already done so.
A. Consider the data modeling steps in Figure 3-1 in the context of FiredUps business.
Without having access to the personnel at FiredUp, it is not possible to plan a data
modeling project. Instead, make a list of questions that you would need to have
answered in order to be able to plan a data modeling project.
The answer here will depend upon the students experience with business subjects. Most of the
questions are answered by the time you complete problem E. A few additional questions that
might be asked would be;
1. Can multiple Stoves be placed on the same Repair order?
2. Do all Customers have an Email address.
B. With regard to requirements determination, consider the sources of requirements in
Figure 3-2. Describe how you think each of these types of requirements might apply to
FiredUp. How would you document the results of your requirements documentation?
First, create a database to store requirements. User Interviews must be done with FiredUp staff
only. You would not be able to interview Customers. You could observe processing of an invoice
as well as observing the processing of a repair. You would need to use the Stove Warranty
registration to obtain some information. The forms and reports would all be manual and probably
no many would be available for this type of firm. At this point, there would be no formally
defined interfaces. You would probably not use domain expertise here. Most of your information
would come from the Robarbs.
C. Consider your answer to FiredUp question E, Chapter 2. How would you go about
validating this model? Make a list of questions that you would ask FiredUp personnel. In
what ways could you use mockups? In what ways could you use prototypes?
The first step is to develop a model based upon your requirements and the business background
you have. Next, sit down with Curt and Julie Robarbs and review the design. At this point, your
primary interest is whether you are missing any data and that you have the relationships and
cardinalities are correct. Ask questions regarding these. You may use mockups of data entry forms
and invoices if the Robarbs resist reviewing ER diagrams. You may want to create a prototype.
On a project this size, it should be fairly easy to create and the prototype would eventually
become the implemented database.

ANSWERS TO TWIGS TREE TRIMMING SERVICE

Review the Twigs Service Project Questions at the end of Chapter 2 and answer the questions
there if you have not already done so.
A. Consider the data modeling steps in Figure 3-1 in the context of Twigs business. Without
having access to Samantha and other personnel (if any) at Twigs, it is not possible to

3-23

Chapter Three Entity-Relationship Data Modeling: Process and Examples

plan a data modeling project. Instead, make a list of questions that you would need to
have answered in order to be able to plan a data modeling project.
The answer here will depend upon the students experience with business subjects. Most of the
questions are answered by the time you complete problem E. A few additional questions that
might be asked would be;
a. Should all jobs be recurring or should.
b. Can there be Services that are neither recurring nor nor-recurring.
c. Is it possible to be considered a customer but not have a service (prospective customer).
B. With regard to requirements determination, consider the sources of requirements in
Figure 3-2. Describe how you think each of these types of requirements might apply to
Twigs. How would you document results of your requirements documentation?
First, create a database to store requirements. User Interviews must be done with Twigs staff. You
could observe Samantha billing a Customer. This may help understand how to relation Customer,
Service, and Chip_Delivery. The forms and reports would all be manual and probably no many
would be available for this type of firm. At this point, there would be no formally defined
interfaces. You would probably not need to use domain expertise here. Most of you information
would be obtained from Samantha.
C. Consider your answer to Twigs question E, Chapter 2. How would you go about
validating this model? Make a list of questions that you would ask Samantha. In what
ways could you use mockups? In what ways could you use prototypes?
The first step is to develop a model based upon your requirements and the business background
you have. Next, sit down with Samantha and review the design. At this point, your primary
interest is whether you are missing any data and that you have the relationships and cardinalities
are correct. Ask questions regarding these. You may use mockups of data entry forms and
invoices if the Samantha resist reviewing ER diagrams. You may want to create a prototype. On a
project this size, it should be fairly easy to create and the prototype would eventually become the
implemented database.

3-24