Relational Model Reading: C&B, Chap 2, 3 & 4. Dept. of Computing Science, University of Aberdeen 2 In this lecture you will learn The concept of Model.

Relational Model

Reading: C&B, Chap 2, 3 & 4

Dept. of Computing Science, University of Aberdeen 2

In this lecture you will learn

• The concept of Model• Simple definitions of terminology in

Relational Model• Why relational model is important?• Initial idea of Relational Querying

Dept. of Computing Science, University of Aberdeen 3

Introduction

• DBMS models data with respect to– Machine

• physical storage of data• involves low level details that are not important for

user

– User• logical view of data (tables)

• We focus on User’s ‘DATA MODEL’

Dept. of Computing Science, University of Aberdeen 4

The Concept of ‘Model’• A model represents ‘something’.

– At a level of detail suitable for its purpose• For example, an architects ‘model’ of a bridge

– Architects model brings the bridge to life even before its construction

• Models can be:– Blueprints for building the real ‘something’s (such as

bridges)– Useful for testing (model a bridge and test it before

construction)– Handy for learning (you start driving lessons on a

simulator)• All our thinking is based on ‘models’ of things

– We build models of everything• Thinking with models is considered ‘SMART’

Dept. of Computing Science, University of Aberdeen 5

Learning with Relational Model

• Theory of databases involves many models– Relational model, ER model

• Relational model is the user’s data model and we learn it first

• DBMSs that use relational model are called Relational Database Management Systems (RDBMS)

• Apply our learning on a few real RDBMS (MSAccess & MYSQL)– Mostly in practical classes– Also in lectures

• Hopefully this training will help us with other RDBMS as well (such as SQL Server and Oracle)

Dept. of Computing Science, University of Aberdeen 6

Why Relational Model?

• Most modern database software is based on relational model– Relational model is what the user sees in a RDBMS

• Relational Model has a simple logical structure (as we will see later)

• It is intuitive (as we will see later)• Relational model has its origins in sound

mathematics– Set theory and theory of relations (hence the name

relational)– We study the mathematics later in the course

• Relational Model describes the pure concepts behind real database systems– Real systems at times may have unfaithful

implementations of these pure concepts

Dept. of Computing Science, University of Aberdeen 7

What is relational model?

• A logical representation of a database• Proposed by IBM’s E. F. Codd in 1970

– Won a Turing Award for the model!!

• Earlier data models were not relational. They were– Hierarchical (tree) or– Network (graph)

• We focus on relational models in this course

Dept. of Computing Science, University of Aberdeen 8

Relational Model is simple

• Relational Database: a collection of connected tables (relations)– Named Tables (relations) have

• Named Columns (attributes)• Unnamed Rows (tuples)

– Domain: a set of allowable values for one or more columns

– Relationships – links or connections– Keys (primary, foreign)– Integrity Constraints

• It’s that simple!

Dept. of Computing Science, University of Aberdeen 9

Example: DreamHome

• DreamHome models data in a typical real estate business– you will use this database in the practical

• DreamHome has data related to– Branch offices– Staff in each branch office– …

• Branch office data– Branch number– Address

• Staff data– Staff number– Staff details

• Since staff belong to a branch office, there is a relationship between staff data and branch office data.– Staff and Branch tables are linked (connected)

Dept. of Computing Science, University of Aberdeen 10

Example: DreamHome 2branchNo Street City Postcode

B005 22 Deer Rd London SW1 4EH

B007 16 Argyll St Aberdeen AB2 3SU

B003 163 Main St Glasgow G11 9QX

B004 32 Clover Dr Bristol BS99 1NZ

B002 56 Clover Dr London NW10 6EU

staffNo fName LName Position Sex DOB Salary branchNo

SL21 John White Manager M 1-Oct-45 30000 B005

SG37 Ann Beech Assistant F 10-Nov-60 12000 B003

SG14 David Ford Supervisor M 24-Mar-58 18000 B003

SA9 Mary Howe Assistant F 19-Feb-70 9000 B007

SG5 Susan Brand Manager F 13-Jun-40 24000 B003

SL41 Julie Lee Assistant F 13-Jun-65 9000 B005

Relation

Relation

AttributesBranch

Staff Primary Key Foreign Key

Dept. of Computing Science, University of Aberdeen 11

Relation (Table)

• Relation has two parts– Schema

• Name of the relation and the set of attributes & their respective domains

• For example, Branch (branchNo:domain1, street: domain2, city:domain3,

postcode:domain4)

– Instance• Set of attributes and their values in a tuple• For example,

(branchNo:B005, street: 22 Dee Street, city:Aberdeen,postcode:AB25 1AB)

Dept. of Computing Science, University of Aberdeen 12

Key

• A key is an attribute or set of attributes that uniquely identify a row

• A table may have many candidate keys• Primary Key: Key selected for unique

identification of rows– For example, branchNo in the Branch table– Cannot be NULL

• Foreign Key: an attribute or a set of attributes in one table that refer to a row in another table– For example, branchNo in Staff table

Dept. of Computing Science, University of Aberdeen 13

Properties of relations

• Rows & columns can be in any order• no two rows are identical• all values belonging to a particular

attribute are from the same domain• attributes are atomic• attributes may have a null value (but not

the primary key)

Dept. of Computing Science, University of Aberdeen 14

Relational model is intuitive

• Many common datasets naturally fit relational model

• Telephone directory– Directory(Name:domain1, Address:domain2,

TelNo:domain3)

• Class Attendance Sheets– Attendance(Name:domain1, IDNo:domain2,

Signature: domain3)

• Try with other datasets you know

Dept. of Computing Science, University of Aberdeen 15

Data Retrieval: Relational Querying

• The main strength of relational model: supports simple yet powerful data retrieval (querying in general)– The whole purpose of modelling data is to be able to

retrieve data items efficiently

• Output of a query is modelled as a relation (table)• In RDBMS users view data always in terms of

tables– if you learn relational model well you can find it easy to

use databases

• SQL (see-quel): a popular query language for relational databases.

Dept. of Computing Science, University of Aberdeen 16

Example: Relational Querying

• Consider the branch relation from the DreamHome database

• Query: we want a list of branches in London

• Output is a sub-relation (sub-table) of the Branch relation (table)

branchNo Street City Postcode

B005 22 Deer Rd London SW1 4EH

B007 16 Argyll St Aberdeen AB2 3SU

B003 163 Main St Glasgow G11 9QX

B004 32 Clover Dr Bristol BS99 1NZ

B002 56 Clover Dr London NW10 6EU

branchNo Street City Postcode

B005 22 Deer Rd London SW1 4EH

B002 56 Clover Dr London NW10 6EU

London Branches

Branch

Dept. of Computing Science, University of Aberdeen 17

Conclusion

• Relational model of database: a collection of connected tables– Conceptually relational model is simple– Querying the data is the main goal of relational

modelling

• User interactions with RDBMS are through SQL (Query language) only

• We now focus on learning SQL– Later we will learn how to organize data into a

relational model (under database design)