DBMS · Lecture notes · Unit 4

Relational Database Design

Database Management Systems (DBMS)

Unit-4

Relational Database

Design

 Looping Outline

  • Functional Dependency
  • Definition and types of FD
  • Armstrong's axioms (inference rules)
  • Closure of FD set
  • Closure of attribute set
  • Canonical cover
  • Decomposition and its types
  • Anomaly in database design and its types
  • Normalization and normal forms
  • 1NF
  • 2NF
  • 3NF
  • BCNF
  • 4NF
  • 5NF

Page 3

What is Functional Dependency (FD)?

Let R be a relation schema having n attributes A1, A2, A3,…, An.

Student
RollNoNameSPIBL
101ravi80
102Mitesh71
103Jay70

Let attributes X and Y are two subsets of attributes of relation R.

If the values of the X component of a tuple uniquely (or functionally) determine the values of the Y component, then there is a functional dependency from X to Y.

This is denoted by X → Y (i.e RollNo → Name, SPI, BL).

It is referred as: Y is functionally dependent on the X or X functionally determines Y.

Diagrammatic representation of Functional Dependency (FD)

X → Y {X1, X2} → Y X → {Y1, Y2} X Y

X1 X2

Y X

Y1 Y2

Example Consider the relation Account(account_no, balance, branch).

account_no can determine balance and branch.

So, there is a functional dependency from account_no to balance and branch.

This can be denoted by account_no → {balance, branch}.

account_no

balance

branch

Types of Functional Dependency (FD)

Full Functional Dependency In a relation, the attribute B is fully functional dependent on A if B is functionally dependent on A, but not on any proper subset of A.

Eg. {Roll_No, Semester, Department_Name} → SPI We need all three {Roll_No, Semester, Department_Name} to find SPI.

Partial Functional Dependency In a relation, the attribute B is partial functional dependent on A if B is functionally dependent on A as well as on any proper subset of A.

If there is some attribute that can be removed from A and the still dependency holds then it is partial functional dependancy.

Eg. {Enrollment_No, Department_Name} → SPI Enrollment_No is sufficient to find SPI, Department_Name is not required to find SPI.

Types of Functional Dependency (FD)

Transitive Functional Dependency In a relation, if attribute(s) A → B and B → C, then A → C (means C is transitively depends on A via B).

Sub_Fac Subject Faculty Age DS Shah DBMS Patel DF Shah

Eg. Subject → Faculty & Faculty → Age then Subject → Age Therefore as per the rule of transitive dependency: Subject → Age should hold, that makes sense because if we know the subject name we can know the faculty’s age.

Types of Functional Dependency (FD)

Trivial Functional Dependency X → Y is trivial FD if Y is a subset of X Eg. {Roll_No, Department_Name, Semester} → Roll_No Nontrivial Functional Dependency X → Y is nontrivial FD if Y is not a subset of X Eg. {Roll_No, Department_Name, Semester} → Student_Name

Page 9

Armstrong's axioms OR Inference rules

Armstrong's axioms are a set of rules used to infer (derive) all the functional dependencies on a relational database.

Reflexivity Augmentation Self-determination If B is a subset of A If A → B If A → A then A → B then AC → BC

Transitivity Pseudo Transitivity Decomposition If A → B and B → C If A → B and BD → C If A → BC then A → B and A → then A → C then AD → C C

Union Composition If A → B and A → C If A → B and C → D then A → BC then AC → BD

Page 11

What is closure of a set of FDs?

Given a set F set of functional dependencies, there are certain other functional dependencies that are logically implied by F.

E.g.: F = {A → B and B → C}, then we can infer that A → C (by transitivity rule)

The set of functional dependencies (FDs) that is logically implied by F is called the closure of F.

It is denoted by F+.

Closure of a set of FDs [Example]

Suppose we are given a relation schema R(A,B,C,G,H,I) and the set of functional dependencies are:

F = (A → B, A → C, CG → H, CG → I, B → H)

  • The functional dependency A → H is logical implied.

We have A → B Transitivity rule A → H B → H

Closure of a set of FDs [Example]

Suppose we are given a relation schema R(A,B,C,G,H,I) and the set of functional dependencies are:

F = (A → B, A → C, CG → H, CG → I, B → H)

  • The functional dependency CG → HI is logical implied.

We have CG → H Union rule CG → HI CG → I

Closure of a set of FDs [Example]

Suppose we are given a relation schema R(A,B,C,G,H,I) and the set of functional dependencies are:

F = (A → B, A → C, CG → H, CG → I, B → H)

  • The functional dependency AG → I is logical implied.

We have A → C Pseudo- AG → I transitivity rule CG → I

Closure of a set of FDs [Example]

Suppose we are given a relation schema R(A,B,C,G,H,I) and the set of functional dependencies are:

F = (A → B, A → C, CG → H, CG → I, B → H)

  • The functional dependency AG → I is logical implied.

We have A → C Augmentation rule AG → CG

AG → CG Transitivity rule AG → I CG → I

Closure of a set of FDs [Example]

Suppose we are given a relation schema R(A,B,C,G,H,I) and the set of functional dependencies are:

F = (A → B, A → C, CG → H, CG → I, B → H)

  • Find out the closure of F.

Several members of F+ are F+ = (A → H, CG → HI, AG → I)

Closure of a set of FDs [Example]

Compute the closure of the following set F of functional dependencies FDs for relational schema R = (A,B,C,D,E,F):

F = (A → B, A → C, CD → E, CD → F, B → E)

  • Find out the closure of F.
A → B & A → CUnion RuleA → BC
CD → E & CD → FUnion RuleCD → EF
A → B & B → ETransitivity RuleA → E
A → C & CD → EPseudo-transitivity RuleAD → E
A → C & CD → FPseudo-transitivity RuleAD → F

F+ = (A → BC, CD → EF, A → E, AD → E, AD → F)

Closure of a set of FDs [Example]

Compute the closure of the following set F of functional dependencies FDs for relational schema R = (A,B,C,D,E):

F = (AB → C, D → AC, D → E )

  • Find out the closure of F.

D → AC Decomposition Rule D → A & D → C D → AC & D → E Union Rule D → ACE

F+ = (D → A, D → C, D → ACE)

Page 20

What is a closure of attribute sets?

Given a set of attributes α, the closure of α under F is the set of attributes that are functionally determined by α under F.

It is denoted by α+.

What is a closure of attribute sets?

Given a set of attributes α, the closure of α under F is the set of attributes that are functionally determined by α under F.

It is denoted by α+.

Algorithm Algorithm to compute α+, the closure of α under F Steps

  1. result = α

  2. while (changes to result) do for each β → γ in F do

  • begin
  • if β ⊆ result then result = result U γ
  • else result = result
  • end

Closure of attribute sets [Example]

Consider the relation schema R = (A, B, C, G, H, I).

For this relation, a set of functional dependencies F can be given as F = {A → B, A → C, CG → H, CG → I, B → H} Step 1.

Find out the closure of (AG)+.

result = α => result = AG Algorithm

A → BA ⊆ AGresult = ABG
A → CA ⊆ ABGresult = ABCG
CG → HCG ⊆ ABCGresult = ABCGH
CG → ICG ⊆ ABCGHresult = ABCGHI
B → H⊆ B ABCGHIresult = ABCGHI

Algorithm to compute α+, the closure of α under F Steps

  1. result = α

  2. while (changes to result) do for each β → γ in F do

  • begin
  • if β ⊆ result then result = result U γ
  • else result = result AG+ = ABCGHI
  • end

Closure of attribute sets [Exercise]

Given functional dependencies (FDs) for relational schema R = (A,B,C,D,E): F = {A → BC, CD → E, B → D, E → A} Find Closure for A Find Closure for CD Find Closure for B Find Closure for BC Find Closure for E Answer A+ = ABCDE CD+ = ABCDE B+ = BD BC+ = ABCDE E+ = ABCDE

Page 25

What is extraneous attributes?

Let us consider a relation R with schema R = (A, B, C) and set of functional dependencies FDs F = { AB → C, A → C }.

In AB → C, B is extraneous attribute. The reason is, there is another FD A → C, which means when A alone can determine C, the use of B is unnecessary (extra).

An attribute of a functional dependency is said to be extraneous if we can remove it without changing the closure of the set of functional dependencies.

What is canonical cover?

A canonical cover of F is a minimal set of functional dependencies equivalent to F, having no redundant dependencies or redundant parts of dependencies.

It is denoted by Fc A canonical cover for F is a set of dependencies Fc such that F logically implies all dependencies in Fc and Fc logically implies all dependencies in F and No functional dependency in Fc contains an extraneous attribute and Each left side of functional dependency in Fc is unique.

F = {A → B, A → C} Decomposition Rule Union Rule Fc = {A → BC}

Algorithm to find canonical cover

Repeat Use the union rule to replace any dependencies in F α1 → β1 and α1 → β2 with α1 → β1β2 Find a functional dependency α → β with an extraneous attribute either in α or in β /* Note: test for extraneous attributes done using Fc, not F */

  • If an extraneous attribute is found, delete it from α → β until F does not change /* Note: Union rule may become applicable after some extraneous attributes have been deleted, so it has to be re-applied */

Canonical cover [Example]

Consider the relation schema R = (A, B, C) with FDs F = {A → BC, B → C, A → B, AB → C} Find canonical cover.

Combine A → BC and A → B into A → BC (Union Rule)

Set is {A → BC, B → C, AB → C} A is extraneous in AB → C Check if the result of deleting A from AB → C is implied by the other dependencies

  • Yes: in fact, B → C is already present Set is {A → BC, B → C} C is extraneous in A → BC Check if A → C is logically implied by A → B and the other dependencies
  • Yes: using transitivity on A → B and B → C.

The canonical cover is: A → B, B → C

Canonical cover [Example]

Consider the relation schema R = (A, B, C, D, E, F) with FDs F = {A → BC, CD → E, B → D, E → A} Find canonical cover.

The left side of each FD in F is unique.

Also none of the attributes in the left side or right side of any of the FDs is extraneous. Therefore the canonical cover Fc is equal to F.

Fc = {A → BC, CD → E, B → D, E → A}

Page 31

What is decomposition?

Decomposition is the process of breaking down given relation into two or more relations. Relation R is replaced by two or more relations in such a way that:

Each new relation contains a subset of the attributes of R Together, they all include all tuples and attributes of R Types of decomposition Lossy decomposition Lossless decomposition (non-loss decomposition)

Lossy decomposition

Customer The decomposition of relation R into R1 and R2 is lossy when the join of R1 and R2 does not yield Ano Balance Bname the same relation as in R.

A01 5000 Rajkot This is also referred as lossy-join decomposition.

A02 5000 Surat The disadvantage of such kind of decomposition Table-1 Table-2 is that some information is lost during retrieval of Ano Balance Balance Bname original relation.

Not Same A01 5000 5000 Rajkot From practical point of view, decomposition A02 5000 5000 Surat should not be lossy decomposition.

Customer Ano Balance Bname A01 5000 Rajkot A01 5000 Surat A02 5000 Rajkot A02 5000 Surat

Lossless decomposition

Customer The decomposition of relation R into R1 and R2 is lossless when the join of R1 and R2 produces the Ano Balance Bname same relation as in R.

A01 5000 Rajkot This is also referred as a non-additive (non-loss)

A02 5000 Surat decomposition.

Table-1 Table-2 All decompositions must be lossless.

Ano Balance Ano Bname Same A01 5000 A01 Rajkot A02 5000 A02 Surat

Customer Ano Balance Bname A01 5000 Rajkot A02 5000 Surat

Page 35

What is an anomaly in database design?

Anomalies are problems that can occur in poorly planned, un-normalized database where all the data are stored in one table.

There are three types of anomalies that can arise in the database because of redundancy are Insert anomaly Delete anomaly Update / Modification anomaly

Insert anomaly

Consider a relation Emp_Dept(EID, Ename, City, DID, Dname, Manager) EID as a primary key

Emp_Dept
EIDEnameCityDIDDnameManager
1RajRajkot1CEShah
2MeetSurat1CEShah
NULLNULLNULL2ITNULL

An insert anomaly occurs when certain attributes cannot be inserted into the database without the presence of another attribute.

Want to insert new department detail (IT)

Suppose a new department (IT) has been started by the organization but initially there is no employee appointed for that department.

We want to insert that department detail in Emp_Dept table.

But the tuple for this department cannot be inserted into this table as the EID will have NULL value, which is not allowed because EID is primary key.

This kind of problem in the relation where some tuple cannot be inserted is known as insert anomaly.

Delete anomaly

Consider a relation Emp_Dept(EID, Ename, City, DID, Dname, Manager) EID as a primary key

Emp_Dept EID Ename City DID Dname Manager A delete anomaly exists when certain attributes are Raj Rajkot CE Shah lost because of the deletion of another attribute. Meet Surat CE Shah Want to delete (Jay)

Jay Baroda IT Dave employee's detail

Now consider there is only one employee in some department (IT) and that employee leaves the organization.

So we need to delete tuple of that employee (Jay).

But in addition to that information about the department also deleted.

This kind of problem in the relation where deletion of some tuples can lead to loss of some other data not intended to be removed is known as delete anomaly.

Update anomaly

Consider a relation Emp_Dept(EID, Ename, City, Dname, Manager) EID as a primary key

Emp_Dept EID Ename City Dname Manager An update anomaly exists when one or more Raj Rajkot CE Sah records (instance) of duplicated data is updated, Meet Surat C.E Shah but not all.

Jay Baroda Computer Shaah Want to update manager Hari Rajkot IT Dave of CE department

Suppose the manager of a (CE) department has changed, this requires that the Manager in all the tuples corresponding to that department must be changed to reflect the new status. If we fail to update all the tuples of given department, then two different records of employee working in the same department might show different Manager lead to inconsistency in the database.

How to deal with insert, delete and update anomaly

Emp_Dept Emp Dept EID Ename City DID Dname Manager EID Ename City DID DID Dname Manager Raj Rajkot CE Shah Raj Rajkot CE Shah Meet Surat C.E Shah Meet Surat IT Dave Jay Baroda IT Dave Jay Baroda 2 EC NULL NULL NULL NULL EC NULL

Such type of anomalies in the database design can be solved by using normalization.

Page 41

What is normalization?

Normalization is the process of removing redundant data from tables to improve data integrity, scalability and storage efficiency.

data integrity (completeness, accuracy and consistency of data)

scalability (ability of a system to continue to function well in a growing amount of work)

storage efficiency (ability to store and manage data that consumes the least amount of space)

What we do in normalization?

Normalization generally involves splitting an existing table into multiple (more than one) tables, which can be re-joined or linked each time a query is issued (executed).

How many normal forms are there?

Normal forms:

1NF (First normal form)

2NF (Second normal form)

3NF (Third normal form)

BCNF (Boyce–Codd normal form)

4NF (Forth normal form)

5NF (Fifth normal form)

As we move from 1NF to 5NF number of tables and complexity increases but redundancy decreases.

Page 44

1NF (First Normal Form)

Conditions for 1NF

Each cells of a table should contain a single value.

A relation R is in first normal form (1NF) if and only if it does not contain any composite attribute or multi-valued attributes or their combinations.

OR A relation R is in first normal form (1NF) if and only if all underlying domains contain atomic values only.

1NF (First Normal Form) [Example - Composite attribute]

Customer
CIDNameAddress
C01raviJamnagar Road, Rajkot
C02MiteshNehru Road, Jamnagar
C03JayC.G Road, Ahmedabad
  • In customer relation address is composite attribute which is further divided into sub-attributes as “Road” and “City”.
  • So customer relation is not in 1NF.

Problem: It is difficult to retrieve the list of customers living in ’Jamnagar’ city from customer table.

The reason is that address attribute is composite attribute which contains road name as well as city name in single cell.

It is possible that city name word is also there in road name.

In our example, ’Jamnagar’ word occurs in both records, in first record it is a part of road name and in second one it is the name of city.

1NF (First Normal Form) [Example - Composite attribute]

Customer Customer CID Name Address CID Name Road City C01 ravi Jamnagar Road, Rajkot C01 ravi Jamnagar Road Rajkot C02 Mitesh Nehru Road, Jamnagar C02 Mitesh Nehru Road Jamnagar C03 Jay C.G Road, Ahmedabad C03 Jay C.G Road Ahmedabad

Solution: Divide composite attributes into number of sub-attributes and insert value in proper sub-attribute.

Exercise Convert below relation into 1NF (First Normal Form)

Person PID Full_Name City P01 ravi Maheshbhai Patel Rajkot

1NF (First Normal Form) [Example - Multivalued attribute]

Student
RnoNameFailedinSubjects
101raviDS, DBMs
102MiteshDBMS, DS
103JayDS, DBMS, DE
104JeetDBMS, DE, DS
105HarshDE, DBMS, DS
106NeelDE, DBMS
  • In student relation FailedinSubjects attribute is a multi- valued attribute which can store more than one values.
  • So above relation is not in 1NF.

Problem: It is difficult to retrieve the list of students failed in ’DBMS’ as well as ’DS’ but not in other subjects from student table.

The reason is that FailedinSubjects attribute is multi-valued attribute so it contains more than one value.

1NF (First Normal Form) [Example - Multivalued attribute]

Student Student Result Rno Name FailedinSubjects Rno Name RID Rno Subject ravi DS, DBMs ravi DS Mitesh DBMS, DS Mitesh DBMS Jay DS, DBMS, DE Jay DBMS Jeet DBMS, DE, DS Jeet DS Harsh DE, DBMS, DS Harsh DS Neel DE, DBMS Neel … … …

Solution: Split the table into two tables in such as way that the first table contains all attributes except multi-valued attribute with same primary key and second table contains multi-valued attribute and place a primary key in it.

insert the primary key of first table in the second table as a foreign key.

Page 50

2NF (Second Normal Form)

Conditions for 2NF It is in 1NF and each table should contain a single primary key.

A relation R is in second normal form (2NF)

if and only if it is in 1NF and every non-primary key attribute is fully dependent on the primary key OR A relation R is in second normal form (2NF)

if and only if it is in 1NF and no any non-primary key attribute is partially dependent on the primary key

2NF (Second Normal Form) [Example]

CIDANOAccessDateBalanceBranchNameFD2
C01 C02A01 A0101-01-2017 01-03-201750000 50000Rajkot RajkotCIDANOAccesssDateBalanceBranchName
C01 C03A02 A0201-05-2017 01-07-201725000 25000Surat SuratFD1

FD1: {CID, ANO} → {AccesssDate, Balance, BranchName} FD2: ANO → {Balance, BranchName} Balance and BranchName are partial dependent on primary key (CID + ANO). So customer relation is not in 2NF.

2NF (Second Normal Form) [Example]

CIDANOAccessDateBalanceBranchNameFD2
C01 C02A01 A0101-01-2017 01-03-201750000 50000Rajkot RajkotCIDANOAccesssDateBalanceBranchName
C01 C03A02 A0201-05-2017 01-07-201725000 25000Surat SuratFD1

Problem: For example, in case of a joint account multiple (more than one) customers have common (one) accounts.

If an account ’A01’ is operated jointly by two customers says ’C01’ and ’C02’ then data values for attributes Balance and BranchName will be duplicated in two different tuples of customers ’C01’ and ’C02’.

2NF (Second Normal Form) [Example]

CustomerTable-1Table-2
CIDANOAccessDateBalanceBranchNameANOBalanceBranchNameCIDANOAccessDate
C01A0101-01-201750000RajkotA0150000RajkotC01A0101-01-2017
C02A0101-03-201750000RajkotA0225000SuratC02A0101-03-2017
C01A0201-05-201725000SuratC01A0201-05-2017
C03A0201-07-201725000SuratC03A0201-07-2017

Solution: Decompose relation in such a way that resultant relations do not have any partial FD. Remove partial dependent attributes from the relation that violets 2NF.

Place them in separate relation along with the prime attribute on which they are fully dependent.

The primary key of new relation will be the attribute on which it is fully dependent.

Keep other attributes same as in that table with the same primary key.

Page 55

3NF (Third Normal Form)

Conditions for 3NF It is in 2NF and there is no transitive dependency.

(Transitive dependency???) A → B & B → C then A → C

A relation R is in third normal form (3NF)

if and only if it is in 2NF and every non-key attribute is non-transitively dependent on the primary key OR A relation R is in third normal form (3NF)

if and only if it is in 2NF and no any non-key attribute is transitively dependent on the primary key

3NF (Third Normal Form) [Example]

Customer ANO BalanceBranchNameBranchAddressFD2
A01 A0250000 40000Rajkot RajkotKalawad road Kalawad RoadANOBalanceBranchNameBranchAddress
A03 A0435000 25000Surat SuratC.G Road C.G RoadFD1

FD1: ANO → {Balance, BranchName, BranchAddress} FD2: BranchName → BranchAddress So AccountNO → BranchAddress (Using Transitivity rule)

BranchAddress is transitive depend on primary key (ANO). So customer relation is not in 3NF.

3NF (Third Normal Form) [Example]

Customer ANO BalanceBranchNameBranchAddressFD2
A01 A0250000 40000Rajkot RajkotKalawad road Kalawad RoadANOBalanceBranchNameBranchAddress
A03 A0435000 25000Surat SuratC.G Road C.G RoadFD1

Problem: In this relation, branch address will be stored repeatedly for each account of the same branch which occupies more space.

3NF (Third Normal Form) [Example]

Customer
ANOBalanceBranchNameBranchAddress
A0150000RajkotKalawad road
A0240000RajkotKalawad Road
A0335000SuratC.G Road
A0425000SuratC.G Road

Table-1 Table-2

BranchNameBranchAddressANOBalanceBranchName
RajkotKalawad roadA0150000Rajkot
SuratC.G RoadA0240000Rajkot
A0335000Surat
A0425000Surat

Solution: Decompose relation in such a way that resultant relations do not have any transitive FD.

Remove transitive dependent attributes from the relation that violets 3NF.

Place them in a new relation along with the non-prime attributes due to which transitive dependency occurred.

The primary key of the new relation will be non-prime attributes due to which transitive dependency occurred.

Keep other attributes same as in the table with same primary key and add prime attributes of other relation into it as a foreign key.

BCNF (Boyce-Codd Normal Form)

Section – 7.4

BCNF (Boyce-Codd Normal Form)

Conditions for BCNF Primary Key Determinant Dependent BCNF is based on the concept of a determinant.

AccountNO → {Balance, Branch}

It is in 3NF and every determinant should be primary key.

A relation R is in Boyce-Codd normal form (BCNF)

if and only if it is in 3NF and for every functional dependency X → Y, X should be the primary key of the table.

OR A relation R is in Boyce-Codd normal form (BCNF)

if and only if it is in 3NF and every prime key attribute is non-transitively dependent on the primary key OR A relation R is in Boyce-Codd normal form (BCNF)

if and only if it is in 3NF and no any prime key attribute is transitively dependent on the primary key

BCNF (Boyce-Codd Normal Form) [Example]

FD1 Student RNO Subject Faculty

  • FD1: RNO, Subject → Faculty DS Patel RNO Subject Faculty
  • FD2: Faculty → Subject
  • So {RNO, Subject} → Subject (Transitivity rule) DBMS Shah DS Jadeja FD2 DBMS Dave In FD2, determinant is Faculty which is not a primary key. So student table is not in DBMS Shah BCNF.

DS Patel Problem: In this relation one student can learn more than one subject with different DBMS Dave faculty then records will be stored repeatedly for each student, language and faculty DS Jadeja combination which occupies more space.

  • Here, one faculty teaches only one subject, but a subject may be taught by more than one faculty.
  • A student can learn a subject from only one faculty.

BCNF (Boyce-Codd Normal Form) [Example]

Student Table-1 Table-2

  • Solution: Decompose relation in RNO Subject Faculty Faculty Subject RNO Faculty such a way that resultant relations DS Patel Patel DS Patel do not have any transitive FD.
  • Remove transitive dependent prime DBMS Shah Shah DBMS Shah attribute from relation that violets Jadeja DS Jadeja Jadeja DS BCNF.

DBMS Dave Dave DBMS Dave

  • Place them in separate new relation along with the non-prime attribute due DBMS Shah Shah to which transitive dependency DS Patel Patel occurred.

DBMS Dave Dave

  • The primary key of new relation will be this non-prime attribute due to which DS Jadeja Jadeja transitive dependency occurred.
  • Keep other attributes same as in that table with same primary key and add a prime attribute of other relation into it as a foreign key.

Multivalued dependency (MVD)

For a dependency X → Y, if for a single value of X, multiple values of Y exists, then the table may have multi-valued dependency.

Student
RNOSubjectFaculty
101DSPatel
101DBMSPatel
101DSShah
101DBMSShah

Multivalued dependency (MVD) is denoted by →→ Multivalued dependency (MVD) is represented as X →→ Y

Page 65

4NF (Forth Normal Form)

Conditions for 4NF A relation R is in fourth normal form (4NF)

if and only if it is in BCNF and has no multivalued dependencies

Student Subject Faculty RNO Subject Faculty RNO Subject RNO Faculty DS Patel DS Patel DBMS Patel DBMS Shah DS Shah DBMS Shah

Above student table has multivalued dependency. So student table is not in 4NF.

Functional dependency & Multivalued dependency

A table can have both functional dependency as well as multi-valued dependency together. RNO → Address RNO →→ Subject RNO →→ Faculty

Student Subject Faculty Address RNO Address Subject Faculty RNO Subject RNO Faculty RNO Address C. G. Road, Rajkot DS Patel DS Patel C. G. Road, Rajkot C. G. Road, Rajkot DBMS Patel DBMS Shah C. G. Road, Rajkot DS Shah C. G. Road, Rajkot DBMS Shah

Page 68

5NF (Fifth Normal Form)

Conditions for 5NF A relation R is in fifth normal form (5NF)

if and only if it is in 4NF and

Student_Result
RIDRNONameSubjectResult
1101RajDBMSPass
2101RajDSPass
3101RajDFPass
4102MeetDBMSPass
5102MeetDSFail
6102MeetDFPass
7103SureshDBMSFail
8103SureshDSPass

Student_Result relation is further decomposed into sub- relations. So the above relation is not in 5NF.

5NF (Fifth Normal Form)

Conditions for 5NF A relation R is in fifth normal form (5NF)

if and only if it is in 4NF and

Student_Result
RIDRNONameSubjectResult
1101RajDBMSPass
2101RajDSPass
3101RajDFPass
4102MeetDBMSPass
5102MeetDSFail
6102MeetDFPass
7103SureshDBMSFail
8103SureshDSPass

Student Result Subject RNO Name RID RNO SID Result SID Name Raj DBMS Pass Meet DS Pass Suresh DF Pass Pass Fail Pass None of the above relations can be further decomposed into sub-relations. So the Fail above database is in 5NF.

Pass

How to find key?

Conditions to find key The attribute is a part of key, if it does not occur on any side of FD The attribute is a part of key, if it occurs on the left-hand side of an FD, but never occurs on the right-hand side The attribute is not a part of key, if it occurs on the right-hand side of an FD, but never occurs on the left- hand side The attribute may be a part of key or not, if it occurs on the both side of an FD

How to find key? [Example]

Let a relation R with attributes ABCD with FDs C → A, B → C. Find keys for relation R. attribute not occur on any side of FDs (D) √ attribute occurs on only left-hand side of an FDs (B) √ attribute occurs on only right-hand side of an FDs (A) X attribute occurs on both the sides of an FDs (C) ?

The core is BD.

B determines C and C determines A, So using transitivity rule B determines A also. So BD is a key.

How to find key? [Exercise]

Let a relation R with attributes ABCD with FDs C → D, C → A and B → C. Find keys for relation R.

The core is B. B determines C which determines A and D, so B is a key. Therefore B is the key.

Let a relation R with attributes ABCD with FDs B → C, D → A. Find keys for relation R.

The core is BD. B determines C and D determines A, so BD is a key. Therefore BD is the key.

Let a relation R with attributes ABCD with FDs A → B, BC → D and A → C. Find keys for relation R.

The core is A. A determines B and C which determine D, so A is a key. Therefore A is the key.

Find (candidate) key & check for normal forms [Example]

Suppose you are given a relation R with four attributes ABCD. For each of the following sets of FDs, do the following: F = (B → C, D → A)

Identify the candidate key(s) for R.

Identify the best normal form that R satisfies (1NF, 2NF, 3NF or BCNF).

Candidate Key is BD

Relation R is in 1NF but not 2NF. In above FDs, there is a partial dependency (As per FD B → C, C depends only on B but Key is BD so C is partial depends on key (BD)) (As per FD D → A, A depends only on D but Key is BD so A is partial depends on key (BD))

Find (candidate) key & check for normal forms [Example]

Suppose you are given a relation R with four attributes ABCD. For each of the following sets of FDs, do the following: F = (C → D, C → A, B → C)

Identify the candidate key(s) for R.

Identify the best normal form that R satisfies (1NF, 2NF, 3NF or BCNF).

Candidate Key is B

Relation R is in 2NF but not 3NF. In above FDs, there is a transitive dependency (As per FDs B → C & C → D then B → D so D is transitive depends on key (B))

(As per FDs B → C & C → A then B → A so A is transitive depends on key (B))

Find (candidate) key & check for normal forms [Example]

Suppose you are given a relation R with four attributes ABCD. For each of the following sets of FDs, do the following: F = (A → B, BC → D, A → C)

Identify the candidate key(s) for R.

Identify the best normal form that R satisfies (1NF, 2NF, 3NF or BCNF).

Candidate Key is A

Relation R is in 2NF but not 3NF. In above FDs, there is a transitive dependency (As per FDs A → B & A → C then A → BC using union rule) and (As per FDs A → BC & BC → D then A → D so D is transitive depends on key (A))

Find (candidate) key & check for normal forms [Example]

Suppose you are given a relation R with four attributes ABCD. For each of the following sets of FDs, do the following: F = (ABC → D, D → A)

Identify the candidate key(s) for R.

Identify the best normal form that R satisfies (1NF, 2NF, 3NF or BCNF).

Candidate Key are ABC & BCD

Relation R is in 3NF but not BCNF.

In the above FDs, both FDs have prime attribute (D and A) in dependent (right) side.

How to normalize database?

A software contract and consultancy firm maintains details of all the various projects in which its employees are currently involved. These details comprise: Employee Number, Employee Name, Date of Birth, Department Code, Department Name, Project Code, Project Description, Project Supervisor.

Assume the following:

Each employee number is unique.

Each department has a single department code.

Each project has a single code and supervisor.

Each employee may work on one or more projects.

Employee names need not necessarily be unique.

Project Code, Project Description and Project Supervisor are repeating fields.

Normalize this data to Third Normal Form.

How to normalize database?

A software contract and consultancy firm maintains details of all the various projects in which its employees are currently involved. These details comprise: Employee Number, Employee Name, Date of Birth, Department Code, Department Name, Project Code, Project Description, Project Supervisor.

UNF

Employee NumberEmployee NameDate of BirthDepartment CodeDepartment NameProject CodeProject DescriptionProject Supervisor
1Raj1-1-851CE1IOTPatel
2Meet4-4-862EC2PHPShah
3Suresh2-2-851CE1IOTPatel
1Raj1-1-851CE2PHPShah

How to normalize database?

UNF

Employee NumberEmployee NameDate of BirthDepartment CodeDepartment NameProject CodeProject DescriptionProject Supervisor
1Raj1-1-851CE1IOTPatel
2Meet4-4-862EC2PHPShah
3Suresh2-2-851CE1IOTPatel
1Raj1-1-851CE2PHPShah

1NF

Employee NumberEmployee NameDate of BirthDepartment CodeDepartment Name
1Raj1-1-851CE
2Meet4-4-862EC
3Suresh2-2-851CE
Employee NumberProject CodeProject DescriptionProject Supervisor
11IOTPatel
22PHPShah
31IOTPatel
12PHPShah

How to normalize database?

1NF

Employee NumberEmployee NameDate of BirthDepartment CodeDepartment Name
1Raj1-1-851CE
2Meet4-4-862EC
3Suresh2-2-851CE
Employee NumberProject CodeProject DescriptionProject Supervisor
11IOTPatel
22PHPShah
31IOTPatel
12PHPShah

2NF

Employee NumberEmployee NameDate of BirthDepartment CodeDepartment Name
1Raj1-1-851CE
2Meet4-4-862EC
3Suresh2-2-851CE
Project CodeProject DescriptionProject Supervisor
1IOTPatel
2PHPShah
Employee NumberProject Code
11
22
31
12

How to normalize database?

3NF

Employee NumberEmployee NameDate of BirthDepartment CodeDepartment CodeDepartment Name
1Raj1-1-8511CE
2Meet4-4-8622EC
3Suresh2-2-851
Project CodeProject DescriptionProject Supervisor
1IOTPatel
2PHPShah
Employee NumberProject Code
11
22
31
12

Questions asked

  1. What is meant by normalization? Write its need. List and discuss various normalization forms.

  2. Consider schema EMPLOYEE(E-ID,E-NAME,E-CITY,E-STATE) and FD = {E-ID → E-NAME, E-ID → E-CITY, E-ID → E-STATE, E-CITY → E-STATE}

  • Find attribute closure for: (E-ID)+
  1. Compute the closure of the following set F of functional dependencies for relation schema R(A, B, C, D, E).

F = { A → BC, CD → E, B → D, E → A}

  • List the candidate keys for R.
  1. Consider schema R = (A, B, C, G, H, I) and the set F of functional dependencies {A → B, A → C, CG → H, CG → I, B → H}. ( Use F+ )
  • Prove that AG → I Holds.

Questions asked

  1. In the BCNF decomposition algorithm, suppose you use a functional dependency α → β to decompose a relation schema r (α, β, γ) into r1 (α, β) and r2 (α, γ).
  • What primary and foreign-key constraint do you expect to hold on the decomposed relations?
  • Give an example of an inconsistency that can arise due to an erroneous update, if the foreign-key constraint were not enforced on the decomposed relations above.
  • When a relation is decomposed into 3NF, what primary and foreign key dependencies would you expect will hold on the decomposed schema?
  1. A college maintains details of its lecturers' subject area skills. These details comprise: Lecturer Number, Lecturer Name, Lecturer Grade, Department Code, Department Name, Subject Code, Subject Name, Subject Level. Assume that each lecturer may teach many subjects but may not belong to more than one department. Subject Code, Subject Name and Subject Level are repeating fields. Normalize this data to Third Normal Form.

Page 85

Page 86

Page 87