Functional Dependencies &

Normalization for Relational DBs

Chapter 7
Contents

1 Introduction
2 Functional dependencies
3 Normalization

2
Contents

1 Introduction
2 Functional dependencies
3 Normalization

3
Overview of Database Miniworld
Design Process
REQUIREMENTS - COLLECTION & ANALYSIS

Functional requirements Data requirements

DBMS–independent

FUNCTIONAL ANALYSIS
CONCEPTUAL DESIGN

High-level transaction Conceptual schema

specification
LOGICAL DESIGN
(DATA MODEL MAPPING)
DBMS–specific

APPLICATION PROGRAM Database schema

DESIGN
PHYSICAL DESIGN

TRANSACTION
Internal schema
IMPLEMENTATION

Application program
4
Application Design Database Design
Introduction
 Each relation schema consists of a number of
attributes and the relational database schema consists
of a number of relation schemas
 Attributes are grouped to form a relation schema
 Need some formal measure of why one grouping of
attributes into a relation schema may be better than
another

5
Introduction
 “Goodness” measures:
 Redundant information in tuples
 Update anomalies: modification, deletion, insertion
 Reducing the NULL values in tuples
 Disallowing the possibility of generating spurious tuples.

6
Introduction
 Redundant information in tuples: the attribute values
belonging to a particular department (DNUMBER,
DNAME, DMGRSSN) are repeated for every employee
who works for that department.

7
Introduction
 Update anomalies: modification, deletion, insertion
 Modification
 As the manager of a department changes we have to update
many values according to employees working for that dept.
 Easy to make the DB inconsistent

8
Introduction
 Insertion:
 How can we create a department before any employees
are assigned to it ??
10
Introduction
 Deletion: if Borg James E. leaves, we delete his tuple
and lose the existing of dept. 1, the name of dept. 1, and
who is the manager of dept. 1
Introduction
 Reducing the NULL values in tuples
 Employees not assigned to any dept.: waste the storage
space
 Other difficulties: aggregation operations (e.g., COUNT,
SUM) and joins

12
Introduction
 Disallowing the possibility of generating spurious tuples

EMP_PROJ (SSN, PNumber, Hours, EName, PName, PLocation)

EMP_LOCS (EName, PLocation)

EMP_PROJ1 (SSN, PNumber, Hours, PName, PLocation)

 Generation of invalid and spurious data during JOINS:

PLocation is the attribute that relates EMP_LOCS and
EMP_PROJ1, and PLocation is neither a primary key nor a
foreign key in either EMP_LOCS or EMP_PROJ1

13
Introduction
 Disallowing the possibility of generating spurious
tuples

14
Introduction
 Disallowing the possibility of generating spurious
tuples

15
Introduction
 Disallowing the possibility of generating spurious
tuples

16
Introduction
 “Goodness” measures:
 Redundant information in tuples
 Update anomalies: modification, deletion, insertion
 Reducing the NULL values in tuples
 Disallowing the possibility of generating spurious tuples
 Normalization

17
Introduction
 Normalization helps DB designers determine the best
relation schemas
 A formal framework for analyzing relation schemas based on
their keys and on the functional dependencies among their
attributes
 A series of normal form tests that can be carried out on
individual relation schemas so that the relational database can
be normalized to any desired degree
 It is based on the concept of normal form 1NF, 2NF, 3NF,
BCNF, 4NF, 5NF
 It is a process which ensures that the data is structured in
such a way that attributes are grouped with the PK.
Attributes that do not directly depend on PK may be
extracted to form a new relation

18
Introduction
 There are two important properties of decompositions:

1) non-additive or losslessness of the corresponding join

2) preservation of the functional dependencies

 Note that property (1) is extremely important and cannot

be sacrificed. Property (2) is less stringent and may be
sacrificed

19
Contents

1 Introduction
2 Functional dependencies
3 Normalization

20
Functional Dependencies (FDs)
 Definition of FDs
 Direct, indirect, partial dependencies
 Inference Rules for FDs
 Equivalence of Sets of FDs
 Minimal Sets of FDs

21
Functional Dependencies (FDs)
 Functional dependencies (FDs) are used to specify
formal measures of the "goodness" of relational
designs
 FDs and keys are used to define normal forms for
relations
 FDs are constraints that are derived from the meaning
and interrelationships of the data attributes
 A set of attributes X functionally determines a set of
attributes Y if the value of X determines a unique value
for Y
X→Y

22
Functional Dependencies (FDs)
 X → Y holds if whenever two tuples have the same value for
X, they must have the same value for Y
 For any two tuples t1 and t2 in any relation instance r(R):
If t1[X]=t2[X], then t1[Y]=t2[Y]
 X → Y in R specifies a constraint on all relation instances r(R)
 Examples:
 social security number determines employee name:
SSN → EName
 project number determines project name and location:
PNumber → {PName, PLocation}
 employee ssn and project number determines the hours per
week that the employee works on the project:
{SSN, PNumber} → Hours

23
Functional Dependencies (FDs)
 If K is a key of R, then K functionally determines all
attributes in R (since we never have two distinct tuples
with t1[K]=t2[K])

24
Functional Dependencies (FDs)
 Definition of FDs
 Direct, indirect, partial dependencies
 Inference Rules for FDs
 Equivalence of Sets of FDs
 Minimal Sets of FDs

25
Functional Dependencies (FDs)
 Direct dependency (fully functional dependency): All
attributes in a R must be fully functionally dependent
on the primary key (or the PK is a determinant of all
attributes in R)
SSN → {Name, BDate, Address, DNO}

EMPLOYEE
SSN Name BDate Address DNO

26
Functional Dependencies (FDs)
 Indirect dependency (transitive dependency): Value
of an attribute is not determined directly by the
primary key
DNO → DName

EMP_DEPT

SSN Name BDate Address DNO DName

27
Functional Dependencies (FDs)
 Partial dependency
 Composite determinant - more than one value is required
to determine the value of another attribute, the
combination of values is called a composite determinant
{SSN, PNumber} in EMP_PROJ
 Partial dependency - if the value of an attribute does not
depend on an entire composite determinant, but only
part of it, the relationship is known as the partial
dependency
SSN → EName , Pnumber → {PName, PLocation}
EMP_PROJ
SSN PNumber Hours EName PName PLocation
28
Functional Dependencies (FDs)
 Definition of FD
 Direct, indirect, partial dependencies
 Inference Rules for FDs
 Equivalence of Sets of FDs
 Minimal Sets of FDs

29
Functional Dependencies (FDs)
 Given a set of FDs F, we can infer additional FDs that
hold whenever the FDs in F hold
 Armstrong's inference rules:
 IR1. (Reflexive) If Y ⊆ X, then X → Y
 IR2. (Augmentation) If X → Y, then XZ → YZ
(Notation: XZ stands for X ⋃ Z)
 IR3. (Transitive) If X → Y and Y → Z, then X → Z
Functional Dependencies (FDs)
 Some additional inference rules that are useful:
 Decomposition: If X -> YZ, then X -> Y and X -> Z
 Union: If X -> Y and X -> Z, then X -> YZ
 Psuedotransitivity: If X -> Y and WY -> Z, then WX -> Z

 The last three inference rules, as well as any other

inference rules, can be deduced from IR1, IR2, and IR3
(completeness property)
Functional Dependencies (FDs)
 Closure of a set F of FDs is the set F+ of all FDs that can
be inferred from F
 Closure of a set of attributes X with respect to F is the
set X+ of all attributes that are functionally determined
by X
 X+ can be calculated by repeatedly applying IR1, IR2,
IR3 using the FDs in F
Exercise
 Consider a relation R(A, B, C, D, E) with the following
dependencies F:
(1) AB → C,
(2) CD → E,
(3) DE → B
 Find {A, B}+ , {A, B, D}+

33
Functional Dependencies (FDs)
 Definition of FD
 Direct, indirect, partial dependencies
 Inference Rules for FDs
 Equivalence of Sets of FDs
 Minimal Sets of FDs

34
Functional Dependencies (FDs)
 Two sets of FDs F and G are equivalent if F+ = G+
 Definition:
 F covers G if G+ ⊆ F+
 F and G are equivalent if F covers G and G covers F
 There is an algorithm for checking equivalence of sets
of FDs
 We can determine whether F covers G by calculating X+
with respect to F for each FD X → Y in G, and then
checking whether this X+ includes the attributes in Y. If
this is the case for every FD in G, then F covers G
 Example:
F = {A → C, AC → D, E → AD, E → H}
G = {A → CD, E → AH}
35
Functional Dependencies (FDs)
 A set of FDs is minimal if it satisfies the following
conditions:
 Every dependency in F has a single attribute for its RHS.
 We cannot remove any dependency from F and have a set
of dependencies that is equivalent to F.
 We cannot replace any dependency X → A in F with a
dependency Y → A, where Y proper-subset-of X ( Y
subset-of X) and still have a set of dependencies that is
equivalent to F
Functional Dependencies (FDs)
 Every set of FDs has an equivalent minimal set
 There can be several equivalent minimal sets
 There is no simple algorithm for computing a minimal
set of FDs that is equivalent to a set F of FDs
 To synthesize a set of relations, we assume that we
start with a set of dependencies that is a minimal set
Contents

1 Introduction
2 Functional dependencies
3 Normalization

38
Normalization
 Normalization: The process of decomposing
unsatisfactory "bad" relations by breaking up their
attributes into smaller relations
 Normal form: Using keys and FDs of a relation to
certify whether a relation schema is in a particular
normal form
 Normalization is carried out in practice so that the
resulting designs are of high quality and meet the
desirable properties
 The database designers need not normalize to the
highest possible normal form (3NF, BCNF or 4NF)

39
Normalization
 Two new concepts:
 A Prime attribute must be a member of some candidate
key
 A Nonprime attribute is not a prime attribute: it is not a
member of any candidate key

40
Normalization
 1NF and dependency problems
 2NF – solves partial dependency
 3NF – solves indirect dependency
 BCNF – well-normalized relations

41
Normalization
 First normal form (1NF): there is only one value at
the intersection of each row and column of a relation -
no set valued attributes in 1NF
→ Disallows composite attributes, multivalued
attributes, and nested relations
 To be part of the formal definition of a relation in the
basic (flat) relational model

42
1NF

43
1NF

44
1NF

45
1NF

46
1NF
EMP_PROJ (SSN, PNumber, Hours, EName, PName, PLocation)

1. SSN, PNumber → Hours

2. SSN → EName
3. PNumber → PName, PLocation

47
1NF
EMP_PROJ (SSN, PNumber, Hours)
EMP (SSN, EName)
PROJ (PNumber, PName, PLocation)

1. SSN, PNUMBER → HOURS

2. SSN → ENAME
3. PNUMBER → PNAME, PLOCATION

48
Normalization
 1NF and dependency problems
 2NF – solves partial dependency
 3NF – solves indirect dependency
 BCNF – well-normalized relations

49
Normalization
 Second normal form (2NF) - all attributes must be fully
functionally dependent on the primary key
 2NF solves partial dependency problem in 1NF
 Method: identify primary keys and group attributes
that relate to the key together to form separate new
relations

50
2NF

51
2NF

➢Problem with 2NF

52
Normalization
 1NF and dependency problems
 2NF – solves partial dependency
 3NF – solves indirect dependency
 BCNF – well-normalized relations

53
Normalization
 A relation schema R is in third normal form (3NF) if it
is in 2NF and no non-prime attribute A in R is
transitively dependent on the primary key

NOTE:
In X → Y and Y → Z, with X as the primary key, we consider
this a problem only if Y is not a candidate key. When Y is a
candidate key, there is no problem with the transitive
dependency .
E.g., Consider EMP (SSN, Emp#, Salary ).
Here, SSN → Emp# →Salary and Emp# is a candidate key

54
Normalization
 3NF solves indirect (transitive) dependencies problem
in 1NF and 2NF
 Method: identify all transitive dependencies and each
transitive dependency will form a new relation, with
non-prime attributes participating in the transitive
dependency and the attribute which determines others
as the attributes for the new relation

55
3NF

56
Exercise
Consider the universal relation R = {A, B, C, D, E, F, G, H, I, J}
and the set of functional dependencies:
1. A, B → C
2. A → D, E
3. B → F
4. F → G, H
5. D → I, J
What is the key for R? Decompose R into 2NF, then 3NF
relations.

57
Normalization
 1NF and dependency problems
 2NF – solves partial dependency
 3NF – solves indirect dependency
 BCNF – well-normalized relations

58
SUMMARY OF NORMAL FORMS based
on Primary Keys

59
General Normal Form Definitions
 The above definitions consider the primary key only
 The following more general definitions take into
account relations with multiple candidate keys

60
General Normal Form Definitions – 2NF
 A relation schema R is in second normal form (2NF) if
every non-prime attribute A in R is not partially dependent
on any key of R

61
General Normal Form Definitions – 2NF

62
General Normal Form Definitions – 3NF
 A relation schema R is in third normal form (3NF) if
whenever a FD X -> A holds in R, then either:
(a) X is a superkey of R, or
(b) A is a prime attribute of R

63
General Normal Form Definitions – 3NF

64
Normalization
 1NF and dependency problems
 2NF – solves partial dependency
 3NF – solves indirect dependency
 BCNF – well-normalized relations

65
Normalization
 A relation schema R is in Boyce-Codd
Normal Form (BCNF) if whenever an FD
X -> A holds in R, then X is a superkey of R

 FD5: Area → County_name

66
 BCNF

Boyce-Codd normal form. (a) BCNF normalization of LOTS1A with the

functional dependency FD2 being lost in the decomposition. (b) A
schematic relation with FDs; it is in 3NF, but not in BCNF.
67
BCNF
 TEACH (Student, Course, Instructor)

68
BCNF
 Three possible pairs

 All three decompositions lose the functional

dependency FD1. The desirable decomposition of those
just shown is 3 because it will not generate spurious
tuples after a join

69
Nonadditive Join Decomposition into BCNF
Schemas

Algorithm 16.5. Relational Decomposition into BCNF with

Nonadditive Join Property
Input: A universal relation R and a set of functional
dependencies F on the attributes of R.
1. Set D := {R} ;
2. While there is a relation schema Q in D that is not in BCNF do
{
choose a relation schema Q in D that is not in BCNF;
find a functional dependency X→Y in Q that violates BCNF;
replace Q in D by two relation schemas (Q – Y) and (X ∪ Y);
};

70
Nonadditive Join Decomposition for Binary
Decompositions
 NJB (Nonadditive Join Test for Binary
Decompositions)
 A decomposition D = {R1, R2} of R has the lossless
(nonadditive) join property with respect to a set of
functional dependencies F on R if and only if either
■ The FD ((R1 ∩ R2) → (R1 − R2)) is in F+, or
■ The FD ((R1 ∩ R2) → (R2 − R1)) is in F+

71
Contents

1 Introduction
2 Functional dependencies
3 Normalization

72
73
Exercise
Consider the universal relation R = {A, B, C, D, E, F, G, H, I,
J} and the set of functional dependencies:
1) A, B → C
2) B, D → E, F
3) A, D → G, H
4) A → I
5) H → J
Decompose R into 2NF, then 3NF relations.

74
Exercise
 Consider the relation:
BOOK (Book_Name, Author, Edition, Year)
 Based on a common-sense understanding of the data,
what are the possible candidate keys of this relation?

75
Key finding algorithms

Extended part
Key and super key
 Superkey of R: A set of attributes SK of R such that no
two tuples in any valid relation instance r(R) will have the
same value for SK. That is, for any distinct tuples t1 and
t2 in r(R), t1[SK] ≠ t2[SK]
 Key of R: A "minimal" superkey; that is, a superkey K
such that removal of any attribute from K results in a set
of attributes that is not a superkey
 If K is a key of R, then K functionally determines all
attributes in R
Key-finding algorithm (1) By Elmasri and Navathe

Input: A relation R and a set of functional dependencies F

on the attributes of R.
Output: a key K of R

1. Set K to contain all attributes in R

2. For each attribute A in K {
compute (K – A)+ with respect to F;
if (K – A)+ contains all attributes in R, then
set K := K – {A}
};
Key-finding algorithm (1) By Elmasri and Navathe

 In algorithm (1), we start by setting K to all the attributes

of R; we then remove one attribute at a time and check
whether the remaining attributes still form a superkey.
 The algorithm (1) determines only one key out of the
possible candidate keys for R; the key returned depends
on the order in which attributes are removed from R in
step 2.
Key-finding algorithm (2)By Hossein Saiedian & Thomas Spencer
Input: A relation R and a set of functional dependencies F on the attributes of R.
Output: all candidate keys of R
Let:
 U contains all attributes of R
 Ul contains attributes of R that occur only on the left-hand side of FDs in F
 Ur contains attributes of R that occur only on the right-hand side of FDs in F
 Ub contains attributes of R that occur on both sides of FDs in F
Note:
 Ul ∩ Ur = ф, Ul ∩ Ub = ф and Ur ∩ Ub = ф
 Ul ∪ Ur ∪ Ub = U
 For every attribute A ∈ U, if A ∈ Ul, then A must be part of every candidate
key of R.
 For every attribute A ∈ U, if A ∈ Ur, then A will not be part of any candidate
key of R.
Key-finding algorithm (2)By Hossein Saiedian & Thomas Spencer
Input: A relation R and a set of functional dependencies F on the
attributes of R.
Output: all candidate keys of R

1. Determine Ul, Ur and Ub

2. If Ul‡+ = U under F, then Ul forms the only key of R and the
algorithm stops here.
Else: move to step 3 // Ul‡+ ≠ U under F
3. Consider every subsets Ubi of Ub: Ubi ⊂ Ub
For each Ubi, if (Ul ∪ Ubi)+ = U under F, then Ki = (Ul ∪ Ubi) is a
candidate key of R (*)

(*) If Ki = (Ul ∪ Ubi) is a candidate key of R, then we need not to check

Ubj ⊂ Ub where Ubi ⊂ Ubj
Key-finding algorithm (2)By Hossein Saiedian & Thomas Spencer
 A simple categorization of attributes into the sets Ul, Ur
and Ub allows to distinguish between those attributes that
will participate in the candidate keys of a relational
database schema and those that do not.
 The algorithm (2) finds all candidate keys.
Exercise 1
Consider the universal relation R = {A, B, C, D, E, F} and
the set of functional dependencies:
1) A → B
2) C, D → A
3) B, C → D
4) A, E → F
5) C, E → D
What is the key for R?
Exercise 2
Consider the universal relation R = {A, B, C, D, E, F} and
the set of functional dependencies:
1) A, D → B
2) A, B → E
3) C → D
4) B → C
5) A, C → F
What is the key for R? Decompose R into 2NF, then 3NF
relations.

85
Exercise 3
Consider the relation R = {A, B, C, D, E, F, G, H, I, J} and the
set of functional dependencies:
1) A, B --> C
2) B, D --> E, F
3) A, D --> G, H
4) A --> I
5) H --> J
What is the key for R?
Exercise 4
Consider the universal relation R = {A, B, C, D, E, F, G, H, I,
J} and the set of functional dependencies:
1) A, B → C
2) A → D, E
3) B → F
4) F → G, H
5) D → I, J

What is the key for R? Decompose R into 2NF, 3NF, and

BCNF relations.

87
Exercise 5
Consider the universal relation R = {A, B, C, D, E, F} and
the set of functional dependencies:
1) A → B
2) C → A, D
3) A, F → C, E
What is the key for R? Decompose R into 2NF, 3NF, and
BCNF relations.
Exercise 6
Consider the following relation books:

BOOK (Book_title, Author_name, Book_type, List_price, Author_affil, Publisher)

Suppose the following dependencies exist:

1) Book_title → Publisher, Book_type
2) Book_type → List_price
3) Author_name → Author_affil

a. What normal form is the relation in?

b. Apply normalization until you cannot decompose the
relations further.

89