Normalization (3NF)
We examined conditions for 2 NF in the last post. Let’s move on to 3 NF, also known as Boyce-Code form.
Here identifying transitive dependency is vital.
Although the table is already in 2 NF but still there is a possibility of anomaly. A person of age 5 can either be in Small or Medium age group. Otherwise, it’s a logical inconsistency.
That’s why usually it’s recommended to keep the database at least in 3 NF form. Higher forms are rarely observed in practice.
Age group actually depends directly on age. The primary key attribute name is not determining the age group.
{ Name } ⤍ { Age } ⤍ { Age group }
This is the transitive dependency and it violates the 3 NF.
3 NF Rule: The data has to be in 2 NF and every attribute should depend only on Key Attribute.
How would a 3 NF look like?
Make sure to remove all the transitive dependencies.
Now data complies with 2 NF as well as 3 NF.
Let’s go through 4 NF & 5 NF the next time.
Writer: Piyush Kulkarni
