What is Neo4j

Neo4j is the world's leading graph database, purpose-built for storing, querying, and analysing highly connected data. Unlike traditional relational databases that use tables and joins, Neo4j stores data as nodes and relationships — making it natural and efficient to work with connected information.

---

A Brief History

• 2007 — Neo4j development begins in Sweden, inspired by the limitations of relational databases for connected data
• 2010 — Neo4j 1.0 released as an open-source graph database
• 2012 — The Cypher query language is introduced, making graph queries accessible
• 2016 — Neo4j 3.0 launches with clustering and the Bolt binary protocol
• 2018 — Neo4j raises $80 million in Series E funding, the largest investment in database history at that time
• 2022 — Neo4j 5.0 released with autonomous clustering and improved Cypher
• Today — Neo4j powers thousands of applications across fraud detection, recommendation engines, knowledge graphs, and more

---

Graph Databases vs Relational Databases

Understanding the difference between graph and relational databases is fundamental:

Relational Databases

In a relational database, data lives in tables connected by foreign keys:

Users Table              Orders Table             Products Table
+---------+------+      +----------+---------+   +-----------+--------+
| user_id | name |      | order_id | user_id |   | product_id| name   |
+---------+------+      +----------+---------+   +-----------+--------+
| 1       | Alice|      | 101      | 1       |   | 501       | Laptop |
| 2       | Bob  |      | 102      | 2       |   | 502       | Phone  |
+---------+------+      +----------+---------+   +-----------+--------+

To find "products purchased by friends of Alice," you need multiple joins across several tables — and performance degrades as data grows.

Graph Databases

In Neo4j, the same data is represented as nodes connected by relationships:

(Alice)-[:FRIENDS_WITH]->(Bob)
(Alice)-[:PURCHASED]->(Laptop)
(Bob)-[:PURCHASED]->(Phone)

Finding "products purchased by friends of Alice" is a natural traversal — no joins required, and performance remains constant regardless of total data size.

---

Key Differences

Feature	Relational Database	Graph Database (Neo4j)
Data model	Tables, rows, columns	Nodes, relationships, properties
Connections	Foreign keys + JOINs	Direct relationships
Query performance on connected data	Degrades with JOINs	Constant time traversal
Schema	Rigid, predefined	Flexible, evolve as needed
Query language	SQL	Cypher
Best for	Structured, tabular data	Connected, relationship-rich data

---

When to Use Neo4j

Neo4j excels when relationships are as important as the data itself:

Use Case	Example
Fraud detection	Identify suspicious transaction rings and shared identities
Recommendation engines	"People who bought X also bought Y"
Knowledge graphs	Connect concepts, documents, and entities for intelligent search
Network and IT operations	Map dependencies between servers, services, and applications
Identity and access management	Model complex permission hierarchies
Supply chain	Track parts, suppliers, and logistics relationships
Social networks	Model friendships, follows, and interactions

---

Neo4j Editions

Edition	Description
Neo4j Community Edition	Free, open-source, single-instance graph database
Neo4j Enterprise Edition	Commercial licence with clustering, security, and advanced features
Neo4j AuraDB	Fully managed cloud graph database (free tier available)
Neo4j AuraDS	Managed graph data science platform in the cloud

---

The Neo4j Ecosystem

Neo4j provides a rich ecosystem of tools:

Tool	Purpose
Neo4j Browser	Web-based interface for running Cypher queries and visualising results
Neo4j Desktop	Local development environment with project management
Neo4j Bloom	Business-friendly graph visualisation and exploration
Neo4j Data Importer	Visual tool for importing CSV data into Neo4j
APOC	Library of hundreds of useful stored procedures and functions
Graph Data Science (GDS)	Library of graph algorithms for analytics
Neo4j Drivers	Official drivers for JavaScript, Python, Java, .NET, and Go

---

The Property Graph Model

Neo4j uses the property graph model, which consists of:

1. Nodes — entities (people, products, places)
2. Relationships — connections between nodes (always directed, always have a type)
3. Properties — key-value pairs on both nodes and relationships
4. Labels — tags on nodes for categorisation (a node can have multiple labels)

(:Person {name: "Alice", age: 30})-[:WORKS_AT {since: 2020}]->(:Company {name: "Acme"})

---

Summary

Neo4j is a graph database designed for connected data. It stores information as nodes and relationships rather than tables and foreign keys, enabling efficient traversal of complex data networks. With its intuitive property graph model, the Cypher query language, and a rich ecosystem of tools, Neo4j is the go-to choice for applications where relationships drive value — from fraud detection to recommendation engines. In the next lesson, we will explore graph data modelling in depth.