• Normalization is a process of decomposing a table into more than one table.
• We need to normalize a table i.e. reduce a table into more tables just to remove insert anomalies, update anomalies.
• It is also used to minimize the redundancy in a table.
• Basically normalization divides larger table into smaller table and establish relationship between them.
• Before moving into normalization depth we must know about functional dependency. Functional dependency states how one column is dependent upon another column.
• While normalizing a table we must keep in mind that the dependency should be preserved.

Types of Normal form

• 1NF
  - In 1NF multi valued and composite attributes are not allowed.
  - For example — suppose we have a table where there is a column of mobile number and a person can have more than one mobile number. Then there will be need to decompose that table in 1NF.
• 2NF
  - The relation should be in 1NF.
  -In 2NF we need to eliminate partial dependency from table.
  - Partial dependency happens when a part of key gives a non key. Suppose AB →C (AB is Key)
  B → C ( B is part of key gives non key element)
  This is known as partial dependency.
  - For lossless decomposition the common attribute should be a key in either of table.
• 3NF
  - The relation should be in 2NF.
  - In 3NF we eliminates the transitive dependency and it is used to remove data duplication.
  - Transitive Dependency occurs when a dependency has no key i.e.
  L.H.S (non-key)→R.H.S (non-key). For example:
  A (Key)→B
  B →C
  C is dependent on B and non of B or C is key.
• BCNF
  - Any relation will be in BCNF if it is in 3NF.
  - For a relation to be in BCNF if for functional dependency the L.H.S is a super key.
  For example : Suppose the dependency are
  AB (Key)→C
  C →B
  So, we must decompose the table in such a way that AC becomes key.