Er Modeling

© Louis D. Nel 1996-2003 102 ER ModelingCOMP 3005 Databases

Module 2

Modeling Data


Object-Oriented Concepts

• Learn the Entity-Relationship approach to modelling data

• Reference: Elmasri & Navathe, 2nd edition chapter 3

Objectives


Object-Oriented Concepts

Topics

• Introduction to E-R model and notation• Example E-R model constructs


Modelling Data

RequirementsCapture

Analysis

ConceptualData Modeling

Customer

(Activity) (Work Product)

Application Requirements

Description ofData Requirements

Enterprise DataModel


Scenarios

• A company wishes to store and access information about its employees, departments and projects.

• Departments will want to generate various reports from the stored data


e.g. Data Description (3.2)

• Company organized into department which have unique name, unique number, manager and date manager started.

• A Department has many locations and controls projects• Each project has a name and unique location• Employee information includes: name, sin number, address,

salary, sex, and birth date. An employee works for one department, but may work on several projects (which could be from different departments)

• Employees have a supervisor and we want to keep track of the number of hours per week they spend on each project

• For insurance purposes the company keeps track of an employees depends. It records the dependents name, sex, birthdate, and relationship to the employee


E-R diagram for Company Data (fig 3.2)

Employee

Dependent

1

1

11

N

N

NM1 N

1

N

Supervisor Supervisee

Bdate

Sin

Name

Sex AddressSalary

Name

NoLocation

StartDate

Supervision

Dependentsof

ManagesDepartment

Location

WorksFor

Project

Controls

Name

No

WorksOn

BdateName Sex Relationship

hours


Entity-Relationship Notation

Manages

Employee Address

Name

Location

Name

Entity

Dependent Weak Entity

Relationship

IdentifyingRelationship(for weak entities)

Attribute

Key Attribute

Discriminator “weak key”

Total participation of E2in R. (E2 can only exist ifit participates in R)

Cardinality 1:N for E1:E2

1 N (min,max)

E must participatein at least min and at most max associations in R

Address Derived Attribute

Multi-valued Attribute

Dependentsof

E1 R E2 E1 R E2 R E


Entities and Attributes

Person

Address

SINDegrees

Entity:a thing with independent existencee.g. person, company, course, job ...

Entity describes a Typevalues are called Entity Sets, orExtensions

Age

Person

Birth Date

Number Street City

First Last

Address

Attribute:Properties of entities, the things whichactually hold the values

Attributes an be assigned null value

Name


Key Attributes

Address

SIN

Key: Attributes whose values can distinguish different entities(No two entity set elements can agree in their key attributes

Person

Number Street CityFirst Last

Name

Some data models require keys, e.g. relational modelother may not e.g. object-oriented model

Name_Address

Key attributes areunderlined inE-R model


Syntactic Description of Entities and Attributes

Address

SINDegrees

Age

Person

Birth Date

Number Street City

First Last

PERSONSIN,Name(First_Name,Last_Name),Address(Number,Street,City),Birth_Date,{Degrees}

( ) = composite attribute{ } = multi-values attribute

Name


Attribute Values and Domains

E

E = EntityA = AttributeA(e) = value of attribute A

of entity type e

V = value set of A in E(Domain of A)

P(V) = powerset of V

A: E -> P(V)

A

Person

E = PersonA = DegreesA(e) = {Bsc, Msc}

V = {Bsc, Msc, PhD}

P(V) = { {} {Bsc} {Msc} {PhD}{Bsc, Msc} {Msc, PhD} {Bsc, PhD}{Bsc, Msc, PhD} }

A(e) ∈ P(V)

Example

Degrees


Initial Entities for Company Database (fig 3.8)

Employee

Bdate

Sin

Name

Sex Address Salary

Name

NoLocation

Manager

Location

Project

Name

No

BdateName Sex

Relationship

Manager Start Date

EmployeeDependent

Department

EmployeeName,SIN,Sex,Address,SalaryBdate,Department,Supervisor,WorksOn{Project, Hours}

Department

Supervisor

WorksOn

Department

DepartmentName,No,{Location}Manager, ManagerStartDate

DependentName,Sex,Bdate,EmployeeRelationship

ProjectName,No,Location,Department


E-R diagram for Company Data (fig 3.2)

Employee

Dependent

1

1

11

N

N

NM1 N

1

N

Supervisor Supervisee

Bdate

Sin

Name

Sex AddressSalary

Name

NoLocation

StartDate

Supervision

Dependentsof

ManagesDepartment

Location

WorksFor

Project

Controls

Name

No

WorksOn


hours


Constraints on Relationship Cardinalities

One to One Many to One

One to Many Many to Many


Employee 11Department

Example Relationships with Constraints

Manages

Employee 1NDepartmentWorks For

Employee MNProjectWorks On

Employee

N

Supervises

1Supervisor

Supervisee


Employee MNProject

Relationships can have attributes

Works On

Number Hours

Employee 11DepartmentManages

StartDate StartDateStartDate

Employee 1NDepartmentWorks For

StartDateStartDate


Employee

Dependent

Dependentsof

Weak Entities

Identifying Entity (owner)Employeename,SIN,address...

Identifying RelationshipEmployeename,SIN,address...

Weak EntityDependentname,age,relationship

has descriminator or partial key

totalparticipationmandatory

In simple cases weak entities can be handled as multivaluedattributes of the owner entity


Company E-R diagram Alternative Notation (fig 3.14)

Employee

Dependent

SupervisorSupervisee

Bdate

Sin

Name

Sex AddressSalary Name

NoLocation

StartDate

Supervision

Dependentsof

ManagesDepartment

Location

WorksFor

Project

Controls

Name

No

WorksOn


hours

(0,1) (1,1)

(1,1)(4,N)

(1,N)

(1,N)

(0,N)

(1,1)(0,N)

(1,1)

(0,1)(0,N)

manageddept.

manager


Supplier Project

Higher Order Relationships

Supply

Part

Quantity ProjNameSupName

PartNo

Supplier ProjectSupplies

Part

StartDateProjNameSupName

PartNoNeedsCanSupply

M N

M

N

M

N


Supplier Project

Equivalent Binary Relationship

Supply

Part

Quantity ProjNameSupName

PartNo

NSupplySS SPj

SPt

Supplier Project

Part

SupName Quantity ProjName

PartNo

N

N

1 1

1


Vehicle

Inheritance

VIN

Car Bicycle

Frame sizeEngine size

Is a

Colour

owner

Number of doors

Component group

Documents

Er Modeling