Oracle 5

NORMALIZATION
Normalization is a process of simplifying the relationship between data elements in a record. Through Normalization a collection of data in a record structure is replaced by successive record structures that are simpler and more predictable and therefore more manageable. It is process of reducing redundancy and eliminating the anomalies. (anamali means inconsistent state.)
Normalization is carried out for four reasons: · To Structure the data so that any pertinent relationships between entities can be represented. · To permit simple retrieval of data in response to query and report requests. · To simplify the maintenance of the data through updates, insertions, and deletion · To reduce the need to restructure or reorganize data when new application requirements arise. · Normalization can improve the quality of design for an application.
There are totally Five Normal Forms(Rules) out of which, any database should necessarily satisfy three of them, those Three Normal Forms are as follows:
First normal form is achieved when all repeating groups are removed so that a record is of fixed length. A repeating group, the re-occurrence of a data item within a record, is actually another relation. Hence, it is removed from the record and treated as an additional record structure, or relation.
Second Normal form should be in first normal form and no partial dependency exists. Means if we have a composite primary key(combination of 2 or more fields) then all the non keys must depend on the composite primary key. no attribute should depend upon the part of key.(for example if Rollno and SSno is a composite key, then no other field in that particular table should be depedent upon either Rollno or ssno individually)
Third Normal form is achieved when Transitive Dependencies are removed from a record design. The general case is as follows : A,B and C are three data items in a record. If C is functionally dependent on B (2nd N.F) and B is functionally dependent on A Then C is functionally dependent on A Therefore, a Transitive Dependency exists. In data management, transitive dependency is a concern because data can inadvertently be lost when the relationship is hidden. In the general case above, if A is deleted, then B and C are deleted also, whether or not this is intended. This problem is eliminated by designing the record for third normal form. Conversion to third normal form removes the transitive dependency by splitting the relation into two separate relations.
BCNF must be 3NF and every determinant is a candidate key. Candidate key is other than primary key, which uniquely identify each row in the table. for example ssno is PK then rollno becomes Candidate key.
4NF and 5NF are depends on instance. we can not give a proper def. bcz it all depends on joins. "CODD's 12 RULES" for a fully Relational DBMS