Nodes, Relationships & Cypher Basics

The Graph Data Model

Nodes (entities) and Relationships (connections) are both first-class, and both can carry properties — a relationship isn't just an implied foreign key, it can hold its own data (e.g. PURCHASED.quantity, FRIENDS_WITH.since). A Label (:Person) categorizes a node roughly like a table name, but a node can carry multiple labels at once, unlike a row belonging to exactly one table.

Cypher — Pattern-Matching Queries

// find a node
MATCH (p:Person {name: "Ada"}) RETURN p

// traverse a relationship — one query, no JOINs
MATCH (a:Person {name: "Ada"})-[:FRIENDS_WITH]->(friend:Person)
RETURN friend

// filter with WHERE, like SQL's WHERE
MATCH (p:Person) WHERE p.age > 30 RETURN p

Cypher's ASCII-art-like syntax ((a)-[:TYPE]->(b)) visually mirrors the graph shape being described — a deliberate readability choice distinct from SQL's row/table-oriented syntax.

Creating Data

CREATE (a:Person {name: "Ada"})

MATCH (a:Person {name: "Ada"}), (b:Person {name: "Bob"})
CREATE (a)-[:FRIENDS_WITH {since: 2020}]->(b)

// MERGE = match-or-create, avoids duplicate nodes on re-run
MERGE (p:Person {email: "ada@example.com"})
  ON CREATE SET p.createdAt = timestamp()

Deleting Nodes with Relationships

// plain DELETE fails if the node still has relationships attached
MATCH (p:Person {name: "Ada"}) DETACH DELETE p
// removes the node AND every relationship connected to it in one step