Lesson 03

The Relational Model

January 5, 2025 1
Entities
 Entity – a person, place, thing, or event
 Entities become our categories of
information
 Customers
 Products
 Orders
 We organize our information in these major
entities (categories)
 These categories become the tables, that
will store our data

2
Relationships
 Relationship – an association between
entities
 Relationships define how our data
files will talk to each other
 Customers “Place” Orders
 Orders “Are For” Products
 There are several key types of
relationships that we will talk about
later on

3
More about Entities
 What does everyone have in common?
 Entities have basic properties too
 Properties of entities are called Attributes
 Attributes describe our categories
 A “Customer” has some basic properties
 Name
 Phone Number
 Address

4
Relational Databases
 Relational Databases
 represent data as a collection of tables
 each row in a table represents a collection of related values

Example

Supplier S

S# SNAME STATUS CITY

S1 Smith 20 London
S2 Jones 10 Paris
S3 Blake 30 Paris
S4 Clark 20 London
S5 Adams 30 Athens

January 5, 2025 5
Relational Model

 Why is it called relational?

 a table is similar to the concept of a relation in mathematics

January 5, 2025 S M Irteza / Dr Rafi Ullah 6

Relation
 Set/Subset
Examples of sets
D1 = {blue; yellow; white} and D2 ={ 1 , 2 }

 Tuple
Examples of tuples
– order of the elements within the tuple is important
< blue, blue > and < yellow, 1 , Paris >

January 5, 2025 S M Irteza / Dr Rafi Ullah 7

Relation
 Consider D1 = {blue; yellow; white} and D2 = {1; 2}
Cartesian Product
 The set of all (binary) tuples such that the first argument
belongs to D1 and the second argument belongs to D2 is
denoted D1xD2.
 A relation on D1xD2 is a collection of tuples < x, y > where all x’s
are taken from D1 and all y’s are taken from D2.

 In other words, a relation on D1xD2 is a subset of D1xD2.

January 5, 2025 8
 Example:
D1 = {blue, yellow, white} and D2 = { 1 , 2}

D1xD2
{ < blue, 1 >,< blue, 2 >,
< yellow, 1 >,< yellow, 2 >,
< white, 1 >,< white, 2 > }

January 5, 2025 9
Relational Model
 Association between tables and relations
 Rows in a table are associated with the tuples in the relation
 Columns in a table are associated with the arguments in the
tuples
 The columns are usually called attributes
 Number of attributes = degree of the relation

January 5, 2025 10
Relational Model – Quick Recap
 Tables (relations)
 Rows (tuples)
 Columns (attributes)
 Unordered tuples
 Atomic attributes
 No duplicate tuples
 Domains

January 5, 2025 11
Relational Data Structure
Supplier S
S# SNAME STATUS CITY
S1 Smith 20 London
S2 Jones 10 Paris
S3 Blake 30 Paris
S4 Clark 20 London
S5 Adams 30 Athens

 Table is the Data Structure for relational model

 Relation = Table
 name of this relation is S.
 Tuple = Row
 This relation has 5 tuples
 Each tuple represents a record of one supplier, so tuples
are seldom called records also.

January 5, 2025 12
Relational Data Structure
Supplier S
S# SNAME STATUS CITY
S1 Smith 20 London
S2 Jones 10 Paris
S3 Blake 30 Paris
S4 Clark 20 London
S5 Adams 30 Athens

 Cardinality = # of rows with data in the relation

 so for S, cardinality = 5.

 Attribute = Columns or fields.

 Degree = # of columns or fields of a relation.

 The relation S has a degree of 4.

January 5, 2025 13
Relational Data Structure
Supplier S
S# SNAME STATUS CITY
S1 Smith 20 London
S2 Jones 10 Paris
S3 Blake 30 Paris
S4 Clark 20 London
S5 Adams 30 Athens

 Domain = A pool of legal values

 For example:
 For S# we have S followed by a positive number.
 For Year we might have a four digit positive number between
1000 and 2003.

 Primary Key = A unique identifier, used to identify one

specific record from among all other records.

January 5, 2025 14
Relations

 A relation consists of 2 parts

 Heading:
 Consists of a fixed set of attributes or columns or fields.
 Defined through the DDL, and usually remains the
same.

 Body
 Consists of the tuples or rows or records.
 Time varying set, i.e., at different intervals of time there
may be different contents in the body of a relation.

January 5, 2025 15
2 Key Properties of Relations
 No Duplicate tuples
 In other words, not all fields are same

 All cells have a single value

 A relation which follows this rule is said to be in the
first normal form.

January 5, 2025 16
Relational Integrity
 The ultimate purpose of a DB is to store
information about the real world.
 It is possible that some information submitted to
the DB does not correspond to any possible
configuration of the real world.
 negative salaries
 impossible dates: 30/02/1999
 negative quantities in stock, etc

January 5, 2025 17
Relational Integrity Rules
- To prevent invalid or illegal values from occurring in the
database we have some rules

 Specific Integrity Rules

 Apply to a specific database, I.e. years may not be
less than zero, and 0 < age < 200 .
 General Integrity Rules
 Apply to all the relational databases.
 Like no two same tuples etc.

January 5, 2025 18
 Candidate Keys
 A relation might have more than one unique identifiers. We
simply choose one as our primary key.

 Attribute K (possibly composite) of relation R is a candidate

key for R , if it satisfies the following two properties.
 UNIQUENESS: no 2 tuples of relation R have the same
values of K.
 MINIMILITY: If K is composite then no subset of K has
the unique property.
- i.e. if you find any column that is unique, don’t add
any more column to the primary key, because this
column by itself is a suitable primary key

January 5, 2025 19
Candidate Keys
 Composite Key Example
 (S#, sname)
 (P#,S#)
 (pname, sname, address, S#)

 Atomic Key Example

 S#
 P#
 sname
 pname

January 5, 2025 20
Primary Keys and Alternate Keys
 A Unique Identifier.
 The primary key is chosen from a set of candidate
keys.
 The primary key is the candidate key chosen to be the
main key for the relation
 The other candidate keys are called alternate keys
 Not duplicable
 NULLS are not allowed in the primary key field.

January 5, 2025 21
Primary Keys
Supplier S

S# SNAME STATUS CITY

S1 Smith 20 London
S2 Jones 10 Paris
S3 Blake 30 Paris
S4 Clark 20 London
S5 Adams 30 Athens

 In the relation/table primary key is “S#”

 As sname, status and city are assumed to be
repeatable or can be duplicate.

January 5, 2025 22
 Primary Keys
 The primary key in SP is (S#, P#) Supplier & Parts SP
 It is composite, as there is no
S# P# QTY
single choice for P.K due to the
format of the table or data. S1 P1 300
 So P.K = (S#, P#) S1 P2 200
S1 P3 400
//composite P.K S1 P4 200
S1 P5 100
 CAUTION: S1 P6 100
 Do not be mistaken by seeing S2 P1 300
S2 P2 400
the data at a time interval, in
S3 P2 200
the table and assuming your S4 P2 200
primary keys. S4 P4 300
S4 P5 400
January 5, 2025 23
Foreign Keys

 A foreign key is an attribute of one

relation R2 whose values are required to
match those values of the primary key of
some relation R1.

January 5, 2025 24
Foreign Keys
 Foreign Keys are used to link data in
two relations. A set of attributes in
the first (referencing) relation is a
foreign key if its value always either:
 Matches a candidate key value in the
second (referenced) relation; or
 Is wholly NULL.
 This is called Referential Integrity

January 5, 2025 25
Foreign Keys: Example
Department: Employee:
DID DName EID EName DID
13 Marketing 15 Yunus Sami 13
14 Accounts 16 Mirza Barki 14
15 Personnel 17 Malik Jahan 13
18 Jan Sher NULL

DID is a Candidate Key for the DID is a Foreign Key in the Employee
Department relation. Each entry relation– each Employee’s DID value is
has a unique value for DID either NULL, or matches an entry in the
Department relation. This links each
employee to (at most) one Department

January 5, 2025 26
Foreign Keys: Example
Student: Enrolment:

ID First Last First Last Course

S01 Ahmad Khan Ahmad Khan CS101
S02 Athar Khalid Ahmad Khan CS222
S03 Sameer Shahid Sameer Shahid CS101
S04 Sameer Khan Sameer Khan CS222

{First, Last} is a Candidate Key for {First, Last} is a Foreign Key in the
the Student relation – no entries Enrolment relation - each {First, Last}
have the same value for both First pair matches exactly one entry in the
and Last. Student relation.

January 5, 2025 S M Irteza / Dr Rafi Ullah 27

Relational Integrity Rules
To prevent invalid or illegal values from occurring in the
database we have some rules

 Specific Integrity Rules

 Apply to a specific database, i.e., Years may not be
less than zero, and 0 < Age < 200 .
 General Integrity Rules
 Apply to all the relational databases.
 Like no two same tuples etc.

January 5, 2025 S M Irteza / Dr Rafi Ullah 28

General Integrity Rules

We have basically two general integrity

rules

1. Entity Integrity Rule.

2. Referential Integrity Rule.

January 5, 2025 S M Irteza / Dr Rafi Ullah 29

1: Entity Integrity Rule

 “No Component of the P.K of a relation is allowed to

accept NULLs”.
 NULL (null value) = unknown or missing value.
 Note that a NULL value is not zero, it is not infinity, or
negative, or FI value. It is simply non existent for the
system.
 P.K = (S#,NULL) is illegal.

January 5, 2025 S M Irteza / Dr Rafi Ullah 30

 2: Referential Integrity Rule
 “The DB must not contain any unmatched foreign key values”.
 e.g., shipment should not have S6 which is not in the suppliers
table as a P.K
 Any state of the DB that doesn't satisfy the integrity rules is
ILLEGAL.
Supplier S Supplier & Parts SP
S# SNAME STATUS CITY S# P# QTY
S1 Smith 20 London S1 P1 300
S2 Jones 10 Paris S1 P2 200
S3 Blake 30 Paris S1 P3 400
S4 Clark 20 London S1 P4 200
S5 Adams 30 Athens S1 P5 100
S1 P6 100
Parts P S2 P1 300
S2 P2 400
P# PNAME COLOR WEIGHT CITY
S3 P2 200
P1 Nut Red 12 London S4 P2 200
P2 Bolt Green 17 Paris S4 P4 300
P3 Screw Blue 17 Rome S4 P5 400
P4 Screw Red 14 London
P5 Cam Blue 12 Paris
P6 Cog Red 19 London

January 5, 2025 S M Irteza / Dr Rafi Ullah 31

How to Avoid Illegal States?
1. Simply reject any operation that would result in an
illegal state, for example:
 if we try to store “S5” in table S with an “S5” entry
existing, then the system should reject that.
2. Accept the operation but perform some additional
compensating operations in order to guarantee the
overall result is still in a legal state.
 E.g. to delete a supplier for which there exists some
shipments, lets say we try to delete “S3” from
supplier, so violation at shipment table occurs as
shipment had “S3” entry as a F.K.

January 5, 2025 S M Irteza / Dr Rafi Ullah 32

End

END OF LESSON 03
(these slides were adapted from original slides of M Ali Shahid)

January 5, 2025 S M Irteza / Dr Rafi Ullah 33