Course contentsShow
← Computer Science
Lesson 1718 of 2,87239. NoSQL and Modern DatabasesFree lesson

Document Database Model

Introduction to document-oriented databases, schema-less design, and how documents differ from relational tables.

Document Database Model

You'll learn what document-oriented databases are and how they fundamentally differ from relational tables.

What is a Document Database?

A document database stores data as documents—typically JSON, BSON, or XML structures—rather than rows in tables. Think of each document as a complete, self-contained record, like a digital filing cabinet where each folder contains all related information about one entity in a single place.

Unlike relational databases that spread information across multiple normalized tables, document databases embrace denormalization: related data lives together in nested structures.

Schema-Less Design

Document databases are often called "schema-less" or "schema-flexible." This means:

  • No predefined structure required: You don't declare columns before inserting data
  • Documents in the same collection can have different fields: One user document might have an address, another might not
  • Easy evolution: Add new fields to new documents without altering existing ones

This flexibility is perfect for applications with evolving requirements or heterogeneous data.

Documents vs Relational Tables

Relational tables store data in rows with fixed columns:

  • Data normalized across multiple related tables
  • Joins required to reassemble complete entities
  • Schema enforced at the database level

Documents store complete entities:

  • Nested objects and arrays inside a single document
  • No joins needed for related data
  • Schema enforced at the application level (if at all)

For example, a user with multiple addresses in a relational DB requires two tables; in a document DB, addresses are simply an array within the user document.

Key Takeaway: Document databases store self-contained, flexible JSON-like documents instead of fixed-schema rows, enabling schema-less design and eliminating the need for joins to retrieve complete entities.