Normalization is the process of efficiently organizing data in a database, and the goal of the normalization process is to eliminating redundant data (storing the same data in many tables) and also ensuring data dependencies are logical and make sense (storing related data in a table and link them based on relationships). Both of these are important since they reduce the amount of space the database can consume and ensure that data is logically stored. As an outcome, this helps reaching the goals such as performance, availability, consistency.
The database experts have suggested adopted a series of guidelines to ensure that the databases are “Normalized”. These are referred as normal forms and are numbered from one to five. In practical applications, and books they are referred as 1NF, 2NF, and 3NF along with the occasional 4NF. Fifth normal form which is very rarely used.
Let’s explore the normal forms.
First normal form (1NF) sets the basic to precede with the other normalization levels, and after 1NF normalization, the tables should meet the following criteria;
Second normal form (2NF) further addresses the concept of removing duplicative data, and after 2NF the tables should meet the following criteria;
Third normal form (3NF) goes one large step further and that should:
Fourth normal form (4NF) has one additional requirement, which can be achieved by restructuring the nature of the table arrangement, and the criteria are:
These normalization guidelines are cumulative. For a database to be in the next level, it must first fulfill all the criteria below that level, and sometimes the database designers deviate from this in order to achieve the business logic.
Detailed steps of normalization will be discussed in the future posts, and check our website regularly for updates.