What is normalization?

Data normalization in the context of databases is the process of organizing data to reduce redundancy and improve data integrity. It is achieved by dividing large tables into smaller ones and linking their relationships.

Main goals:

• Eliminating data duplication.
• Improving data integrity (avoiding insertion, update, and deletion anomalies).
• Enhancing storage and query efficiency (in some cases).

The normalization process is based on the principles of normal forms (NF). The most common are:

• First Normal Form (1NF): Atomicity of data in cells, absence of repeating groups of columns.
• Second Normal Form (2NF): Meets 1NF and each non-key attribute is fully functionally dependent on the entire primary key.
• Third Normal Form (3NF): Meets 2NF and there are no transitive functional dependencies of non-key attributes on the primary key through other non-key attributes.

Example:

Original table "Orders_Products":

Order ID	Order Date	Product ID	Product Name	Product Price	Quantity	Customer Name	Customer Address
1	2023-10-27	101	Phone	500	1	Ivan Petrov	Pushkin St., 10
1	2023-10-27	102	Headphones	100	2	Ivan Petrov	Pushkin St., 10
2	2023-10-27	101	Phone	500	1	Maria Ivanova	Lermontov St., 5

After normalization (up to 3NF):

"Orders" table:

Order ID	Order Date	Customer ID
1	2023-10-27	1
2	2023-10-27	2

"Order Items" table:

Item ID	Order ID	Product ID	Quantity
1	1	101	1
2	1	102	2
3	2	101	1

"Products" table:

Product ID	Product Name	Product Price
101	Phone	500
102	Headphones	100

"Customers" table:

Customer ID	Customer Name	Customer Address
1	Ivan Petrov	Pushkin St., 10
2	Maria Ivanova	Lermontov St., 5

However, excessive normalization can lead to an increased number of joins when executing queries, which may reduce performance. Therefore, database design often seeks a balance between normalization level and performance. Denormalization is the process of intentionally violating normal forms to improve performance, usually by adding redundancy.