🔗Edge

NEW

Also known as: Link, Arc

A connection between nodes in a graph that represents a relationship and may carry direction, weight, or properties.

DESCRIPTION

Edges connect nodes to encode relationships such as friendship, purchase, or dependency. They can be directed or undirected, and may have weights, labels, and properties in property graphs. Edge multiplicity enables multigraphs and parallel relationships. Edge-centric metrics like betweenness and weights influence pathfinding and flow. Modeling edge types carefully is crucial for query performance and semantics in graph databases.

ALTERNATIVE DEFINITIONS

Wikipedia

In graph theory, an edge is a connection between two vertices.

Minimal mathematical definition focusing on connectivity.

Source

Apache TinkerPop

An edge connects two vertices and can have a label and properties in a property graph.

Extends the definition to labeled property graph semantics.

Source

Neo4j

Relationships (edges) connect nodes and always have a direction, a type, and can have properties.

Stresses directionality and typing in Neo4j’s model.

Source

RELATED TERMS

Graph Property Graph Adjacency Directed Graph Weighted Graph Relationship Type Triple Path Multigraph

COMPETING TERMS

Node

KNOWLEDGE AREAS

Semantic & Knowledge Modeling Data Modeling AI Data Architecture Data Science

TARGET AUDIENCE

Data Modeler Data Architect AI Engineer Data Scientist Software Developer

EXTERNAL SOURCES

• Wikipedia - Edge (graph theory)
• Apache TinkerPop - Definition of edges in graphs
• Neo4j - Relationships in property graphs
• NetworkX - Edges API examples

TOP QUESTIONS

What is the difference between directed and undirected Edges?

Directed edges have an orientation (source→target); undirected edges represent symmetric relationships.

Category: Concepts

When should Edges carry properties?

When relationships need attributes like timestamps, weights, roles, or provenance for richer queries.

Category: Practice

How do weighted Edges change analysis?

Weights affect shortest paths, centrality, and flow, encoding strength or cost of relationships.

Category: Analysis

What are common mistakes when modeling Edges?

Encoding relationship attributes as node properties instead of edge properties, losing query expressiveness.

Category: Mistakes

Can multiple Edges connect the same two Nodes?

Yes; multigraphs allow parallel edges representing different relationship types or instances.

Category: Concepts

How does Edge direction affect graph queries?

Direction constrains traversal and path semantics, influencing performance and correctness.

Category: Operations

How are Edges indexed in graph databases?

By relationship type and key properties for efficient traversal and filtering.

Category: Operations

What is an Edge label or type?

A categorization (e.g., PURCHASED, FRIEND_OF) that conveys semantics and supports constraints.

Category: Concepts

How do Edges factor into GNNs?

Edges define message-passing routes and may contribute features like weights or types.

Category: AI