Database Normalisation 101
Database Normalisation 101
Database Normalisation
When altering an entity in a table with redundancies, you must modify all
repeated instances of information, and any other information related to the
changed data. Otherwise, the database becomes inconsistent and anomalies
happen when making changes.
1. Insert anomaly - When trying to insert a new employee in the finance sector,
you must also know the manager's name. Otherwise, you cannot insert data into
the table.
3. Delete anomaly - If Joshua decides to leave the company, deleting the row
also removes the information that a finance sector exists.
First Normal Form (1NF) Repeating and complex values split up,
making all instances atomic.
The general steps in database normalization work for every database. The
specific steps of dividing the table as well as whether to go past 3NF depend on
the use case.
To rework the database table into the 1NF, values within a single field
must be atomic. All complex entities in the table divide into new rows or
columns.
The second normal form in database normalization states that each row in
the database table must depend on the primary key.
The table splits into two tables to satisfy the normal form:
Section 1: Database Normalisation Page |6
The resulting database in the second normal form is currently two tables with no
partial dependencies.
The database is currently in third normal form with three relations in total. At this
point, the database is normalized. Any further normalization steps depend on the
use case of the data.