Graph Database

Definition: A graph database stores data as nodes (things) and relationships (connections between things) instead of mainly using tables and rows.
It is especially useful when the relationships between data are important, such as people connected to friends, products, or transactions. (Neo4j Graph Intelligence Platform)

Simple example

Imagine a social network:

(Alice) ──FRIENDS_WITH──> (Bob)
   │                         │
   │ FOLLOWS                 │ FOLLOWS
   ▼                         ▼
(Charlie) <────────────── (David)

Here:

• Nodes = Alice, Bob, Charlie, David

• Relationships = FRIENDS_WITH, FOLLOWS

• Properties = things like name, age, location

The important thing is that the connections themselves are stored as part of the database, making it easy to traverse from one entity to another. (Neo4j Graph Intelligence Platform)

3 examples

1. Social network:
   Alice → FRIENDS_WITH → Bob → FOLLOWS → Charlie
   You can quickly find connections between users.

2. Recommendation system:
   User → BOUGHT → Product → BELONGS_TO → Category
   You can find products related to things a user already bought.

3. Fraud detection:
   Person → OWNS → Account → USED_AT → Device
   If many suspicious accounts are connected to the same device, the relationships can reveal a possible fraud network.

Common example

Neo4j is a popular graph database. Its basic model is:

Node ── Relationship ── Node

For example:

(Reyhaneh) ──WORKS_AT──> (Company)

Neo4j uses this property graph model, where both nodes and relationships can contain properties. (Neo4j Graph Intelligence Platform)

Easy way to remember:

Relational DB: “What data do I have?” → Tables
Document DB: “What does this object look like?” → Documents
Key-Value DB: “I know the key; give me the value.”
Graph DB: “How are these things connected?”

---

Database My-Journey-In-Codeless