Data Normalization - Information Systems Reengineering, Integration and Normalization - page 338

UNF

1NF 2NF 3NF BCN 4NF

5NF

Fig. 9.1 Levels of relational data normal forms

UNF

1NF

Normalization S#

S#

S-Name S-Name

Enroll

Enroll

Multiple values

Atomic values

{Math}
{Chem}

Brown {Math}

S1

S1

Brown

S1 Brown (Chem)

Fig. 9.2 Normalization of first normal form (1 NF)

dependency relates data with each other in certain constraint rules. The UNF can be
normalized up to fifth normal form as shown below (Fig. 9.1 ):
Data redundancy means that the elimination of redundant data is without loss
of information. An unnormalized relation implies that too many data relationship
dependency in the same table, which is difficult to operate. Normalized database is
easy-to-use for operations of insert, update, and delete.
Normalization starts with UNF with a universe of relations such that each table
may have multiple values. For example, the following table is an UNF with its mul-
tiple valued attributes as shown below (Fig. 9.2 ):
Where relation R (S#, S-Name, {Enrol}) is an UNF with { } as repeating group
data
Functional dependency implies that a determinant can determine dependency
fields. Inclusion dependency (ID) implies that a subclass data must be within its