Разделы презентаций


Normalization Theory

Содержание

Monday, August 12, 2019CONTENTSWhat is the purpose of the normalization theory of RMBad DB projectsFunctional dependencies Multivalued dependencies Join dependenciesNormal forms Design of relational model schema

Слайды и текст этой презентации

Слайд 1Lecture 8. Normalization theory of the relational model
National Aviation University
Computer

Science Faculty
Department of Software Engineering

Lecture 8. Normalization theory of the relational modelNational Aviation UniversityComputer Science FacultyDepartment of Software Engineering

Слайд 2Monday, August 12, 2019
CONTENTS
What is the purpose of the normalization

theory of RM

Bad DB projects

Functional dependencies
Multivalued dependencies
Join dependencies

Normal forms
Design of

relational model schema

Monday, August 12, 2019CONTENTSWhat is the purpose of the normalization theory of RMBad DB projectsFunctional dependencies Multivalued

Слайд 3Monday, August 12, 2019
What is the purpose of the normalization

theory
The theory of relational model normalization establish :
how initial

relational schema may be transformed into other relational schema, which
equivalent initial one in some sense and
Is better it in any sense.
Thus this theory should answer to the following questions:

What criteria of equivalence of relational schemas exist?;
What criteria of estimation of relational schemas quality exist?
What techniques of equivalent transformations of relational schemas exist?
Monday, August 12, 2019What is the purpose of the normalization theoryThe theory of relational model normalization establish

Слайд 4Monday, August 12, 2019

Bad DB design (1)


Customers
We have set a

limit of 9 purchases. What if a customer has more

than 9 purchases?
What to do if purchases are less 9? Set values to NULL? What to do if it is necessary to delete purchase in the middle of the list?
What we have to do if it is necessary to order cusomers’ orders.
How does search condition look like? For example to find customer that buy parts with No 2?: (PN1 = 2) OR (PN2 = 2) OR (PN3 = 2) ... OR (PN9 = 2)

9 purchases




Monday, August 12, 2019Bad DB design (1)CustomersWe have set a limit of 9 purchases. What if a

Слайд 5Monday, August 12, 2019
Bad DB design (2)
Insertion anomaly: Data cannot

be added because some other data is absent .
Update anomaly:

Data inconsistency or loss of data integrity can arise from data redundancy/repetition and partial update .
Deletion anomaly: Data maybe unintentionally lost through the deletion of other data .

Why! It is possible when one relation contains information about two ore more entities of the application domain


CUSTOMER-PURCHASE

Monday, August 12, 2019Bad DB design (2)Insertion anomaly: Data cannot be added because some other data is

Слайд 6Monday, August 12, 2019
Normalization



Normalization is a step by step reversible

process of equivalent transformation of one relational schema into other

that has better characteristics. Every step of such transformation is called normal form.

Compound (not atomic) values - 1NF
Not full (partial) functional dependence - 2NF
Transitive functional dependence - 3NF
Multivalued dependence - 4NF
Join dependence - 5NF

There are the following unwanted properties of relations and normal forms that remove corresponding properties:

Monday, August 12, 2019NormalizationNormalization is a step by step reversible process of equivalent transformation of one relational

Слайд 7Monday, August 12, 2019
Compound domains and the First Normal Form

(1NF)


Relation is in the first normal form (1NF) if

all it’s attributes are based on atomic (simple) domains and consequently the values in table cells are simple (atomic).
Relation is called normalized if it in first normal form .




Monday, August 12, 2019Compound domains and  the First Normal Form (1NF) Relation is in the first

Слайд 8Monday, August 12, 2019
Functional dependencies (FD)
Let’s relation R with attributes

A and B is given. In relation R attribute B

functionally depends on attribute A or A functionally determines B, if and only if every value of the projection R[A] is linked exactly to one value of the projection R[B]. Such functional dependence is denoted as R.А → R.В.
The set of attributes A is called determinant for the set of attributes B.


Formally FD is defined in such a way :

The presence of FD is property of the relational schema, but not of instance of relational schema, and reflects semantics of the AD.

The set of FD can be viewed as a set of integrity constraints on the relation scheme; it should be preserved under decomposition.

Monday, August 12, 2019Functional dependencies (FD)Let’s relation R with attributes A and B is given. In relation

Слайд 9Monday, August 12, 2019
Keys
Set of attributes K in relation R

is candidate key of the relation R if :
each attribute

of the relation R functionally depends on K;
any attribute in K cannot be removed from K without violation of property (a).

Assertion: Any relation has candidate key.

Set of attributes K in relation R is called superkey of the relation R if each attribute of the relation R functionally depends on K.

Monday, August 12, 2019KeysSet of attributes K in relation R is candidate key of the relation R

Слайд 10Monday, August 12, 2019

Properties of functional dependencies
Properties 1), 2), 3)

are Armstrong axioms
Armstrong’s axioms are a
sound and complete set of

inference rules
Monday, August 12, 2019Properties of functional dependenciesProperties  1), 2), 3) are Armstrong axiomsArmstrong’s axioms are asound

Слайд 11Monday, August 12, 2019
Logical inference of functional dependencies
Let’s R

have set of functional dependencies F and the dependence А

→ С that is not in F. Dependence А → С is logically implied by (or logically deduced from) F if it may be inferred from F with the help of functional dependencies axioms.

For example, if we have the relation R(A, B, C) and F contains the dependence А → С, then the following dependences are logically implied from F:

(А, С) → В - continuation property is applied;
(А, С) → (В, С) - augmentation property is applied .

Monday, August 12, 2019Logical inference of functional dependencies Let’s R have set of functional dependencies F and

Слайд 12Monday, August 12, 2019
Closure, completeness, equivalence and minimal cover of

FD
Let’s relation R have set of functional dependencies F. The

set of all functional dependencies logically implied by F is called (logical) closure of F. It is notified as F+. It is obvious, that F ⊆ F+ и F+ = F++.

Set of functional dependencies F is complete if F = F+.

Two sets of dependencies F and G are (logically) equivalent if F+ = G+.

Lets given sets of functional dependencies F and G such that G ⊂ F. G is a cover of F if G+ =F+. If G is minimal then G is called basis of dependencies of F or minimal cover.
NOTE: Minimal cover isn't necessarily unique.

Monday, August 12, 2019Closure, completeness, equivalence  and minimal cover of FDLet’s relation R have set of

Слайд 13Monday, August 12, 2019
FD и сущности предметной области
Thesis. If application

domain contains functional dependence А → В there exists class

of the entities that consist of attributes (A, B). More over in this class set of attributes A is an unique identifier of entities of this class (key) and B are properties of these entities .

If А → В1, А → В2, …, А → Вn, the exists class of entities with attributes(А, В1,…,Вn), where А – unique identifier and В1,…, Вn – are ordinary attributes.

This thesis gives formal basis for identifying entities in AD.

Monday, August 12, 2019FD и сущности предметной областиThesis. 	If application domain contains functional dependence  А →

Слайд 14Monday, August 12, 2019
Not full (partial) functional dependencies and second normal

form (2NF)
Let’s given relation with schema R(A, B, C). Functional

dependence R.A → R.B is full if B does not depends functionally from any С ⊂ А that does not contained in B.

CUSTOMER-PURCHASE

Attribute Q-TY depends fully from (PN, CN, DATE)
Attributes NAME and CITY depends fully from CN
Attributes NAME and CITY depends not fully (partially) from (PN, CN, DATE)

Monday, August 12, 2019Not full (partial) functional dependencies and second normal form (2NF)Let’s given relation with schema

Слайд 15Monday, August 12, 2019
Anomalies of insertion, deleting and updating when

not full FD exist
CUSTOMER-PURCHASE
Update anomalies. When changing customer city it

is necessary to remember that information about customers may be duplicated.
Inset anomalies. When it is necessary to enter information about new customer (Ann) we may do it only when it will do purchase.
Delete anomalies. On deleting information about purchase of Ann we have to delete information about this customer.
Monday, August 12, 2019Anomalies of insertion, deleting and updating  when not full FD existCUSTOMER-PURCHASEUpdate anomalies. When

Слайд 16Monday, August 12, 2019
The second normal form (2NF)
Relation is

in the second normal form (2NF) if it is in

the first normal form and all its nonprimary attributes are depend fully from candidate key .

Теорема (Хита). The relation R with attributes А, В, С , where R.A → R.B, is equal to natural join of the projections R[A, B] and R[A, C].

Algorithm of reduction to 2NF. Given relation R with set of attributes M. If in R there is not full functional dependence R.A → R.B of non primary attribute B from a candidate key A, the relation R is decomposed into the following two relations: R [A, B] and R [M - B]. If the resulting relations are still not in the second normal form, the mentioned algorithm is applied to these relations again.

Such splitting is called binary decomposition.

Monday, August 12, 2019The second normal form (2NF) Relation is in the second normal form (2NF) if

Слайд 17Monday, August 12, 2019
Example of reduction to the 2NF

CUSTOMER-PURCHASE
CUSTOMER
PURCHASE


Monday, August 12, 2019Example of reduction to the 2NFCUSTOMER-PURCHASECUSTOMERPURCHASE

Слайд 18Monday, August 12, 2019
Example of reduction to the 2NF -

Summary
Source relation contains information from 2 entities, every resulting relations

contain information about one entity each.
Resulting relations do not contain anomalies of deletion, inserting and updating.
Source relation can be restored from resulting relations with the help of natural join.
Such decomposition do not lose functional dependencies. They may be restored from decomposed relations.
Monday, August 12, 2019Example of reduction to the 2NF - SummarySource relation contains information from 2 entities,

Слайд 19Monday, August 12, 2019

Transitive dependencies and the Third Normal Form

(3NF)
1) Condition В → А is necessary in order to

exclude trivial transitive dependence like this:



Student card No

Tax ID

Student name

2) Conditions С ⊄ В, В ⊄ А are necessary to exclude the following trivial transitive dependencies:



А



С

В




В

А


С

Monday, August 12, 2019Transitive dependencies  and the Third Normal Form (3NF)1) 	Condition В → А is

Слайд 20Monday, August 12, 2019





Anomalies of insertion, deleting and updating when

transitive FD exist
DEPARTMENT-FACULTY


DEPARTMENT entities
FACULTY entities
Availability transitive dependencies in a relation

means the relation contains information from more that one entity.
As a result such relations imply anomalies of insertion, deletion, updating.
Monday, August 12, 2019Anomalies of insertion, deleting and updating  when transitive FD existDEPARTMENT-FACULTYDEPARTMENT entitiesFACULTY entitiesAvailability transitive

Слайд 21Monday, August 12, 2019
The Third Normal Form (3NF)
The relation is

in the third normal form (3NF) if it is in

the second normal form and does not contain transitive dependencies of nonprimary attributes from candidate keys .
Other words all nonprimary attributes must functionally depend ONLY from candidate keys.

Algorithm of the relation reduction to 3NF. Let’s given the relation R with attributes A, B, C and there are functional dependencies R.A → R.B and R.В → R.С. The relation R decomposed into following two relations: R[A, B] and R[B, С]. If the resulting relations are still not in the third normal form, the mentioned algorithm is applied to these relations again.

Monday, August 12, 2019The Third Normal Form (3NF)The relation is in the third normal form (3NF) if

Слайд 22Monday, August 12, 2019
Example of reduction to the 3NF
DEPARTMENT
FACULTY


Monday, August 12, 2019Example of reduction to the 3NF DEPARTMENTFACULTY

Слайд 23Monday, August 12, 2019
Example of reduction to the 3NF -

Summary
Results the same as in reduction to the 2NF:
Source relation

contains information from 2 entities, every resulting relations contain information about one entity each.
Resulting relations do not contain anomalies of deletion, inserting and updating.
Source relation can be restored from resulting relations with the help of natural join.
Such decomposition do not lose functional dependencies. They may be restored from decomposed relations.
Monday, August 12, 2019Example of reduction to the 3NF - SummaryResults the same as in reduction to

Слайд 24Monday, August 12, 2019
Strong 3NF (S3NF)
Note that 3NF requires

absence of transitive dependence of nonprimary attributes but not all

attributes of the relation . Strong 3NF requires absence of transitive dependence of ALL attributes of a relation

Relation is in strong 3NF, if it is in 3NF and does not contain transitive dependencies of ALL attributes from candidate keys .

This relation is in the 3NF, but still contains information about two entities. So it hold anomalies.

Monday, August 12, 2019Strong 3NF (S3NF) Note that 3NF requires absence of transitive dependence of nonprimary attributes

Слайд 25Monday, August 12, 2019

Reduction to the S3NF
Algorithm of reduction

to the S3NF is the same as for 3NF


Student

Subject
Teacher
STUDY






Student

Teacher

Teacher

Subject
TEACHING-WHOM
TEACHING-WHAT
NOTE. One of

the functional dependence is lost!!!
Monday, August 12, 2019Reduction to the S3NF Algorithm of reduction to the S3NF is the same as

Слайд 26Monday, August 12, 2019
Boyce-Codd normal form (BCNF)
Relation R is in

Boyce-Codd normal, if every its determinants is a superkey.
Assertion. S3NF and

BCNF are equivalent
Monday, August 12, 2019Boyce-Codd normal form (BCNF)Relation R is in Boyce-Codd normal, if every its determinants is

Слайд 27Monday, August 12, 2019
Multivalued dependencies and the Fourth Normal Form

(4NF)
Thesis: If in an application domain there is no

direct relationship between attributes A and B, and it is necessary to fix such relationship in one relation, the only correct decision is to determine, that all values of attribute A are related to all values of attribute B, and vise versa.

TEACHING

Monday, August 12, 2019Multivalued dependencies and  the Fourth Normal Form (4NF) Thesis: If in an application

Слайд 28Monday, August 12, 2019
Definition of the multivalued dependency (MVD)
Given relation

R with attributes (set of attributes) А, В, С. There

exists multivalued dependency В of А (or А determines В multivalued), denoted as А →→ В, if for given set of values attributes from A there exist set of related values attributes of B and this set of B-values does not depends from values of attributes C.

Example: In the relation TEACHING the are the following MVD:

Let’s there is relation R(A,B). The MVDs А →→ ∅ and А →→ В are called trivial because they exist in any relations.

Subject →→ Teacher Subject →→ Bood

Monday, August 12, 2019Definition of the multivalued dependency  (MVD)Given relation R with attributes (set of attributes)

Слайд 29Monday, August 12, 2019
MVD axioms
Given relation R with attributes (set

of attributes) А, В, С.
Multivalued dependences have the following axioms

:

1) Complementation axiom

If А →→ В, then А →→ С

2) Augmentation axiom

If А →→ В and V ⊆ W, then (А, W) →→ (В, V)

3) Transitivity axiom

If А →→ В and В →→ С, then А →→ С – В

Monday, August 12, 2019MVD axiomsGiven relation R with attributes (set of attributes) А, В, С.Multivalued dependences have

Слайд 30Monday, August 12, 2019
Axioms that relates FD и MVD
1) Replication axiom


The following two axioms relates functional and multivalued dependencies .
If

А → В, then А →→ В

2) Coalescence axiom

If А →→ В and Z ⊆ B, and for some W, that is not intersect with B we have W → Z, then A → Z

Monday, August 12, 2019Axioms that relates FD и MVD1)	Replication axiom 	The following two axioms relates functional and

Слайд 31Monday, August 12, 2019
Some additional properties of MVD
1) Union
If А

→→ В and А →→ С, then А →→ (В,

С)

If А →→ В and (W, В) →→ Z ,
then (W, А) →→ Z – (W, В)

2) Pseudo-transitivity

3) Mixed pseudo-tranisitivity

If А →→ В and (А,В) →→ С, then А →→ (С - В)

4) Intersection and difference

If А →→ В and А →→ С, then А →→ В ∩ С, А →→ В – С, А →→ С – В

Monday, August 12, 2019Some additional properties of MVD1) UnionIf А →→ В and А →→ С, then

Слайд 32Monday, August 12, 2019
The fourth nornal form (4NF)
The relation R

is in fourth normal form (4NF), if from existence of

nontrivial multivalued dependence X →→ Y (where Y ⊄ Х) it is follows that Х is a superkey of the relation R.

Assertion. Lets relation R consists of attributes (set of attributes) А, В, С. Dependence А →→ В exist in R if and only if R = R[A, B] * R[A, C].

Monday, August 12, 2019The fourth nornal form (4NF)The relation R is in fourth normal form (4NF), if

Слайд 33Monday, August 12, 2019
Reduction to the 4NF and embedded MVD
Algorithm

reduction to the 4NF. Lets given relation R with attributes

(set of attributes) А, В, С, and given multivalued dependence R.A →→ R.B. Relation R decomposed into the following two relations: R[A, B] и R[B, С].

If resulting relations are not in 4NF the algorithm is applied once more to these relations .

Multivalued dependency is embeded if it is absent in the relation but exists in the projection of the relation by some attributes.

Monday, August 12, 2019Reduction to the 4NF and embedded MVDAlgorithm reduction to the 4NF. Lets given relation

Слайд 34Monday, August 12, 2019
Join dependency (JD) and the Fifth Normal

Form (5NF)
Lets R is a relation with attributes (set of

attributes) A1, A2,,.., An. Relation R have join dependency with respect of A1, A2,,.., An, that is denoted as *(A1, A2,…, An), if relation R is equal to natural joins of all of its projections over A1, A2,…, An:
R = πA1(R)*πA2(R)*... * πAn(R) ⇔ R = R[A1] * R[A2]*…*R[An])


JD is trivial if one of Ai is equal to the list of all attributes of the relation R.

A join dependency is implied by the candidate keys of R if each of Ai (1 ≤ i ≤ n) are superkeys of R.

Monday, August 12, 2019Join dependency (JD) and  the Fifth Normal Form (5NF)Lets R is a relation

Слайд 35Monday, August 12, 2019
Relationships between JD and MVD
Every JD of

the form *(A, B) in relation with schema R(A,B), where

А and В - set of attributes, is equivalent to the MVDs А ∩ В →→ А and А ∩ В →→ В. (Any MVD is JD, but not wise versa!!!)
But there exist JD that are not equivalent any MVD. An example of such JD in relation R(A, B, C) is the dependency *((A,B), (B, C), (A,C)). It is not equivalent to any MVD. Example:

In the example to the left relation contains JD *((A,B), (B, C), (A,C)). It may be verified by calculating: πA1(R) * πA2(R) *... * πAn(R) .
But it does not contain any nontrivial MVD.

It may be convinced by testing that no one of the following multi-valued dependency exist:
A →→B, A →→C, B →→A, B →→C, C →→A,C →→B.

Monday, August 12, 2019Relationships between JD and MVDEvery JD of the form *(A, B) in relation with

Слайд 36Monday, August 12, 2019
The Fifth Normal Form - 5NF
Relation R

is in the fifth normal form (5NF) if and only

if for all of its nontrivial JD *(А1, А2,…, Аn) all sets of attributes Ai are superkeys of R.

This normal form is also called project-join normal form (PJNF).

Assertion. Because of any multivalued dependency is also join dependency, any relation in PJNF (5NF) is also in 4NF .

Classic example to motivate 5NF involves a join n-way decomposition that cannot be derived by a sequence of 2-way decompositions

Monday, August 12, 2019The Fifth Normal Form - 5NFRelation R is in the fifth normal form (5NF)

Слайд 37Monday, August 12, 2019
Example of the relation in the 5NF
Let’s

the is a relation TBS(TCH, BOK, SBJ), where we record

information aboutsuch ssues :

what teaches what books are used;

what books in what subjects are used;

what subjects by what teachers are taught.

The fact that the relation contains the following information:

Reznichenko uses in his lectures the book «SQL language»,

The book «SQL language» is used in the subject DB&KB» and

Reznihcenko has lectures by subject DB&KB.

Does not mean that «Reznichenko uses the book ”SQL Language” in his lectures by subject DB&KB»

Relation TBS is in 5NF because it do not have JDs.

Monday, August 12, 2019Example of the relation in the 5NFLet’s the is a relation TBS(TCH, BOK, SBJ),

Слайд 38Monday, August 12, 2019
Example of the relation that violates 5NF,

and reduction it to the 5NF
If relation TBS has additional

rule (as a business rule of the application domain):

then relation TBS has JD *((TCH, BOK), (BOK, SBJ), (TCH, SBJ)) and this relation is not in 5NF because it has the only candidate key that coincide with all attributes of the relation, that is (TCH, BOK, SUBJ).

In this case relation TBS reduces to 5NF in such a way:

«From the facts:
- teacher t uses in his lectures book b,
- book b is used in subject s and
- teacher t has lectures by subject s
follows that teacher t uses book b in lectures by subject s»,

TBS(TCH, BOK, SBJ) ⇒TB(TCH,BOK), BS(BOK, SBJ), TS(TCH, SBJ)

Monday, August 12, 2019Example of the relation that violates 5NF,  and reduction it to the 5NFIf

Слайд 39Monday, August 12, 2019
Example of the relation that violates 4NF,

and reduction it to the 4NF
If relation TBS has additional

rule (as a business rule of the application domain):

Then relation TBS has JD *((TCH, BOK), (TCH, SBJ)) or it is the same as the relation has the following MVDs TCH →→ BOK, TCH →→ SBJ, and this relation neither in 5NF nor in 4NF.
In this case TBS reduces to the 4NF (and more over to the 5NF) in such a way:

«From the facts:
- teacher t uses in his lectures book b,
- teacher t has lectures by subject s
follows that teacher t uses book b in lectures by subject s»,

TBS(TCH, BOK, SBJ) ⇒TB(TCH,BOK), TS(TCH, SBJ)

Monday, August 12, 2019Example of the relation that violates 4NF,  and reduction it to the 4NFIf

Слайд 40Monday, August 12, 2019
Design of relational model schema
Formal description of

the relational schema design task
Decomposition of the relational schema
Equivalence

of relations
Loosless decomposition with data preservation
Loosless decomposition with dependencies preservation
Equivalence of the normal forms
Criteria of a relation qualities
Monday, August 12, 2019Design of relational model schemaFormal description of the relational schema design task  Decomposition

Слайд 41Monday, August 12, 2019
Formal definition of the relational schema design

task
In this definition it is necessary to clarify the following

items:
what procedure must be used to convert one set of relations into another;
what does equivalence of the two schemas mean ;
how can we estimate that one relational schema is better than another .

Thesis of universal relation. All application domain may be represented as one universal relation that contains all attributes of the domain.

Formal definition of the design task. Lets given relational schema S0, that contains schema of the only (universal) relation R:
S0 = {R = }, where U – set of attributes, а G – set of dependencies, It is necessary to find equivalent relational schema SD, represented as the set of relations R1,…, Rn:
SD = {Ri = , i = 1, 2, ..., n},
that should be better in any sense that schema S0.

Monday, August 12, 2019Formal definition of the relational schema  design taskIn this definition it is necessary

Слайд 42Monday, August 12, 2019
Decomposition of the relational schema
It is said

that decomposition has the property of looseless join, if R

is a natural join of the relations R1, R2,…, Rn., that is R = R1 * R2 *…* Rn

Decomposition is the only operation that is used while splitting relational schemas

Decomposition relation R(M) with set of attributes M into the set of relations R1, R2,…, Rn with attributes M1, MA2,…, Mn is a procedure that satisfy the following conditions:
М1 ∪ М2 ∪ … ∪ Мn = М. That is any attribute of R belongs at least one of relation Ri and all attributes of Ri should be defined in R .
All relations Ri, i = 1, 2,..., n, are projections of the relation R over attributes that are contained in the Ri, that is Ri(Mi) = πMi(R)

Monday, August 12, 2019Decomposition of the relational schemaIt is said that decomposition has the property of looseless

Слайд 43Monday, August 12, 2019
Equivalence of relational schemas by dependencies
Equivalence by

dependencies. Two sets of the relations are equivalent by dependencies,

if they defined on the same set of attributes and they have the same set of dependencies (functional and multivalued).


Formally, lets given two schemas S0 and SD, that was defined previously. They are equivalent by dependencies if:



Where U, Ui are attributes of the schemas S0 and SD: and G, Gi are dependencies of S0 and SD.

Monday, August 12, 2019Equivalence of relational schemas  by dependenciesEquivalence by dependencies. Two sets of the relations

Слайд 44Monday, August 12, 2019
Equivalence of relational schemas by data
Equivalence by

data. Two sets of relations equivalent by data if natural

join of both set of relations gives the identical relations.

If source and resulting schemas are S0 and SD, equivalence by data means that splitting of the relation is done by using loosless decomposition.
How does loosless decomposition may be achived?

Assretion. If R1(U1) R2(U2) are decomposition of R(U) that preserve functional and/or multivalued dependencies, then this decomposition provides lossless join if and only if: U1 ∩ U2 → (or →→) U1 – U2
OR
U1 ∩ U2 → (or →→) U2 – U1


Monday, August 12, 2019Equivalence of relational schemas  by dataEquivalence by data. Two sets of relations equivalent

Слайд 45Monday, August 12, 2019
Equivalence of the normal forms
Property of loosless

join not always guarantee dependency preservation.
At the same time

relation splitting that provides dependency preservation not always guantee the property of loosless join.

Equivalence of the normal forms.
Decomposition of the universal relation up to the 3NF preserve equivalence by data and dependencies.
Converting universal relation to the BCNF preserve equivalence by data but not preserve equivalence by dependencies.

Monday, August 12, 2019Equivalence of the normal formsProperty of loosless join not always guarantee dependency preservation. At

Слайд 46Monday, August 12, 2019
Criteria of the relational schema quality
Let

us consider how can we estimate the schema quality: that

is criteria that one schema is better than other.

One schema is better that other if it does not have data manipulation anomalies .

Actually quality of the schema may be estimated by normal forms. The higher form is used the more qualified schema is received .
Monday, August 12, 2019Criteria of the relational schema quality Let us consider how can we estimate the

Обратная связь

Если не удалось найти и скачать доклад-презентацию, Вы можете заказать его на нашем сайте. Мы постараемся найти нужный Вам материал и отправим по электронной почте. Не стесняйтесь обращаться к нам, если у вас возникли вопросы или пожелания:

Email: Нажмите что бы посмотреть 

Что такое TheSlide.ru?

Это сайт презентации, докладов, проектов в PowerPoint. Здесь удобно  хранить и делиться своими презентациями с другими пользователями.


Для правообладателей

Яндекс.Метрика