What is Normalization of the Data?

Third Normal Form (3NF) is a level of database normalization aimed at eliminating transitive dependencies in a relational schema. A relation is in 3NF if it is already in Second Normal Form (2NF) and if, for every non-trivial functional dependency
$X \rightarrow A$

, either
$X$

is a superkey, or
$A$

A is a prime attribute (an attribute that is part of some candidate key). This condition ensures that all non-prime attributes are only dependent on superkeys and not on other non-prime attributes, thereby reducing redundancy and improving data integrity.

Transitive dependencies occur when a non-prime attribute depends on another non-prime attribute rather than directly on a candidate key. For instance, in a relation where StudentID → DeptID and DeptID → DeptName, the attribute DeptName is transitively dependent on StudentID through DeptID. This violates 3NF because DeptName is not directly dependent on the key. Decomposing such relations into smaller, related tables—each focused on a single subject—eliminates these transitive dependencies and brings the schema into compliance with 3NF.

The data is in Third Normal form in that any query of when a client is booked cannot display a duplicate. Thus, to avoid data duplication, we have provided the CustomerID in the Appointment table. There will not be any duplication of appointment details, since at the same time we cannot have more than one appointment at a time for a particular customer.

In laymen terms; Third Normal Form (3NF) is a way of organizing data in a database so that everything is neat, efficient, and not repeated unnecessarily. Think of it like organizing your contacts or a library. If you have a list of books, and each book includes the author’s name and email, as well as the publisher’s name and address, you might end up repeating the same author and publisher info over and over again. This is not only wasteful, but also risky—if the author changes their email, you’d have to update it in many places. 3NF solves this by saying: only include information that directly relates to the main subject—in this case, the book. Any information that really belongs to someone or something else (like the author or publisher) should be moved into its own separate list or table. That way, each piece of information is stored just once, making it easier to update, less error-prone, and more organized.