Normalization

Introduction

Normalization is the process of organizing data to remove redundancy and dependency anomalies. The first three normal forms (1NF, 2NF, 3NF) cover 99% of real-world design.

Understanding the topic

Core concepts to understand:

• 1NF: atomic values — no lists or repeating groups in a column.
• 2NF: 1NF + no partial dependency on a composite PK.
• 3NF: 2NF + no transitive dependencies (non-key columns depend only on PK).
• BCNF is a stricter version of 3NF.
• Goal: each fact stored exactly once, in the right place.

Syntax reference

Visual workflow / architecture:

Bad (denormalized)
orders(id, customer_name, customer_email, items 'A,B,C')
    ▲ duplicates name/email everywhere
    ▲ items as CSV breaks queries

Good (3NF)
users(id, name, email)
orders(id, user_id, ...)
order_items(order_id, product_id, qty, price)

Real-world use

Every healthy production schema is at least in 3NF. Reporting warehouses sometimes denormalize for speed — that's a separate trade-off.

Best practices

• Aim for 3NF in OLTP (transactional) databases.
• Denormalize only with measured reason and clear ownership.
• Each fact lives in exactly one place.

Hands-on exercise

Interview preparation — practice these questions:

• Q1. State 1NF, 2NF, 3NF.
• Q2. Example of a 1NF violation.
• Q3. Example of a transitive dependency.
• Q4. When to denormalize?
• Q5. What is BCNF?