Normalization process is process to obtain relationship scheme property’s that is nicely becomes higher normal form so that conditions under this fulfilled:
- loop of data which is not necessarily. Redundancy cannot be eliminated at all because good for integrity referential, but redundancy in optimalisation can. For data amounts which are not too much might not too influential in the case of usage hard disk. But imagined if there is thousands, even millions of redundancy, possibly will hardly having an in with space usage.
- Eliminates anomaly. Anomaly basically is un-consistency (inconsistency). Takes example there are commutation of name of from Bank “Maju” becomes Bank “Maju Terus” counted 4 records. If commutation of name of only done at one of record only, hence happened un-consistency that is one relationship bank numbers with 2 name of different bank.
Decomposition tables of can lessen the redundancy and eliminates anomaly.
Scheme through normalization process has advantages as follows:
a. minimization of Storage measure needed to by data storage.
b. minimization of Risk inconsistency data at data base.
c. Minimization possibility that renewal anomaly.
d. Maximizes data structures stability.
Normal form
Purpose of normalization process is convert relationship to become higher normal form. There is having immeasurable level of normal form, that is :
a. First normal form (1NF)
b. Second normal form (2NF)
c. Third normal form (3NF)
d. Normal Form Boyce-Codd (BCNF)
e. Fourth normal form ( 4NF)
f. Fifth normal form ( 5NF)
First normal form to eliminate plural valuable attribute. Second normal form to eliminate partial dependence. Third normal form to eliminate transitive dependence. Normal Form Boyce-Codd to eliminate anomaly remains caused [by] functional dependence. Fourth normal form to eliminate plural value dependence. Fifth normal form to eliminate anomaly remains.
Three normal forms firstly relates to functional dependence. Meanwhile fifth and fourth form related to redundancy caused by dependence many values ( multi-valued dependencies).
First Normal Form
First normal form is equivalent with definition of model relational. Relationship is first normal form (1NF) if all its(the attribute values was simple ( non composite).
Second Normal Form
Having second normal form is rule must have is in the form of first normal ( 1NF), and all attributes is not main must hinge is functional fullness at relationship lock.
Relationship at second normal form storage should not facts about part of relationship lock. Eliminating second normal form is parsial dependence and still having anomaly’s that is practically in acceptable.
Third Normal Form
Normal form third has rule must have is in the form of normal second ( 2NF) and relationship may not load functional dependence among attributes is not main.
Normal form third eliminates transitive dependence, initially normal form third is thought as a form of normal of top / very end. But then detectable stronger normal form that is normal form Boyce-Codd.
Normal Form Boyce-Codd ( BCNF)
BCNF has rule that is each main attribute hinged is functional fullness at each lock where the lock is non part of his (its. Relationship is BCNF (optimal) if every relationship attributes determinant was locked relationship. Relationship is BCNF (optimal) if whenever facts is kept about some attributes, hence this attributes is one relationship locks. BCNF can have lock multiple. Important property of BCNF is relationship doesn't have information which redundancy.
Fourth Normal Form
Relationship in the form of fourth normal (4NF) if relationship in BCNF and didn't contain dependence many values. To eliminate dependence many values out of one relationship, we divide relationship to become two new relationships. Every relationship contains two attributes having relationship many values.
Fifth Normal Form
Fifth normal form (5NF) deals with property so-called join without existence of loss of information (lossless join). Fifth normal form (5NF) also is called as PJNF (normal projection-join of form). This Cases very rare emerges and difficult to be detected practically.
.jpg)
