Relational Model E.F. Codd at IBM 1970

Chapter 3 (ed. 7 – Chap. 5)

Relational Model

• “A relational model of data for large shared data banks”

• Most popular• simplest • most uniform data structures • most formal• files + mathematical foundation

• Example of a information in a database?

EMPLOYEE FNAME MINIT LNAME SSN BDATE ADDRESS SEX SALARY SUPERSSN DNO

  John  B Smith 123456789 1965-01-09 731 Fondren, Houston, TX M 30000 333445555 5

Franklin  T Wong 333445555 1955-12-08 638 Voss, Houston, TX M 40000 888665555 5

Alicia  J Zelaya 999887777 1968-01-19 3321 Castle, Spring, TX F 25000 987654321 4

Jennifer  S Wallace 987654321 1941-06-20 291 Berry, Bellaire, TX F 43000 888665555 4

Ramesh  K Narayan 666884444 1962-09-15 975 Fire Oak, Humble, TX M 38000 333445555 5

Joyce  A English 453453453 1972-07-31 5631 Rice, Houston, TX F 25000 333445555 5

Ahmad  V Jabbar 987987987 1969-03-29 980 Dallas, Houston, TX M 25000 987654321 4

James  E Borg 888665555 1937-11-10 450 Stone, Houston, TX M 55000 null 1

DEPARTMENT DNAME DNUMBER MGRSSN MGRSTARTDATE

  Research 5 333445555 1988-05-22

Administration 4 987654321 1995-01-01

Headquarters 1 888665555 1981-06-19

Database

• Object, instance, entity• Relationships

• The relational model captures all of this in a table

Relational Model

• a table is composed of rows and columns• row - a collection of related data values

– describing an entity or relationship instance

• column – same attribute for different entities

• Relation is a table of values• Database is a collection of relations

Glossary of terms:

• Table -> RELATION• Row -> TUPLE• Column header -> ATTRIBUTE• Data type of a column -> DOMAIN

Glossary Cont’d

• DOMAIN D– specify data type, format for each domain– is of data-type (or format)– set of atomic values (relational model

requirement)• TUPLE

– no composite values (all values are atomic)– no multivalued attributes (First Normal Form)

Relations

• RELATION Characteristics – Set of tuples - not ordered– Values within tuples are ordered

• ATTRIBUTES– Ai is an attribute with a domain dom(Ai)– degree of a relation - the number of attributes

• RELATION SCHEMA: R(A1, A2,...,An):– includes relation name R and list (set) of attributes Ai

DB Schema – 2 notations

• EMPLOYEE(Fname, Minit, Lname, SSN, Bdate, Address, Sex, Salary)

• EMPLOYEE Fname Minit Lname SSN Bdate Address Sex Salary

Relations

• RELATION (instance) r( R )– current relation state – set of n-tuples (where n is a number of attributes)– r = {t1, t2,...,tm} each n-tuple t is an ordered list on n values

– t = <v1, v2,...vn> where vi is an element of dom(Ai) or null

– tj[Ai] is the value vi for attribute Ai of tuple tj

– the tuples themselves are not ordered

r( R ) is a subset of what? (use domains of attributes)

dom(A1) X dom(A2) X...X dom(An)

Relations

• relation intension refers to the schema• relation extension to the state• all tuples within a relation are distinct

Constraints

1. DOMAIN CONSTRAINTS:– 1st normal form 1NF

• value of each attribute must be an atomic (single) value from the domain for that attribute

• no composite attributes

Constraints

2. KEY CONSTRAINTS:– All elements are distinct

• no two tuples can have the same combination of values for all their values (uniqueness constraint). Therefore, all elements are distinct.

• How can this be enforced?• superkey sk

– any set of attributes with property: no two tuples from the relation have the same combination of

values for those attributes– t1[sk] != t2[sk]

– every relation has at least one superkey, what is it?

Constraints cont’d

• key is a minimal superkey if – we cannot remove any attribute and still have uniqueness

constraint hold– (city, state, zip)– Tuscaloosa has > 1 zip, same zip for Cottondale, Coaling

• candidate key is any one of the keys• primary key PK is designated candidate key

– underlined in both relation and ER models

• EMPLOYEE(Fname, Minit, Lname, SSN, Bdate, Address, Sex, Salary)

Constraints

3. ENTITY INTEGRITY CONSTRAINTS:– no primary key value can be null

Referential Constraint - Relationships

4. REFERENTIAL INTEGRITY CONSTRAINTS– a tuple in one relation that refers to another relation must

refer to an existing tuple in that relation

Student(Name, SSN, HomePhone, Address, OfficePhone, Age, GPA)Department(Name, DeptHead, Location, College)

• Students have a major• To Student relation add?

Referential Constraint - Relationships

Student(Name, SSN, HomePhone, Address, OfficePhone, Age, GPA, Major)Department(Name, DeptHead, Location, College)

Constraints

4. REFERENTIAL INTEGRITY CONSTRAINTS– a tuple in one relation that refers to another relation

must refer to an existing tuple in that relation• specified between two relations• maintain the consistency among tuples in two relations• if one relation refers another, then the primary key of the referred

relation is a foreign key FK in the referring relation. t1[PK] = t2[FK]

• Can a foreign key can be null?• Yes, for example if an employee does not have a supervisor

Referential Constraint - Relationships

Student(Name, SSN, HomePhone, Address, OfficePhone, Age, GPA, Major)

Department(Name, DeptHead, Location, College)

Referential Constraint

Name SSN HomePhone Address OfficePhone Age GPA Major

Name DeptHead Location College

Relations

• RELATIONAL DATABASE SCHEMA– a set of relation schemas and a set of referential integrity

constraints S = {R1, R2,..., Rm} and a set of integrity constraints IC

• RELATIONAL DATABASE INSTANCE DB– DB = {r1, r2,...rm}