benedekrozemberczki/karateclub

Health v2 · maintenance only

Adoption not part of the score

2286 stars · 255 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Karate Club is a Python library of unsupervised machine learning algorithms for graph-structured data, built as an extension of NetworkX. It provides node and graph embedding techniques plus overlapping and non-overlapping community detection methods with a scikit-learn-like API.

Use cases

• embed nodes of a networkx graph with node2vec or deepwalk
• detect communities in a network using louvain or label propagation
• compute graph-level embeddings for graph classification
• cluster a graph with unsupervised learning methods
• do small-scale graph mining research in python
• learn node representations for downstream machine learning tasks

When to choose

• you work with NetworkX graphs and want a consistent sklearn-style API for graph ML
• you need many community detection or graph embedding algorithms in one library
• you are doing small-scale graph mining or network science research

When to avoid

• you need large-scale, distributed graph processing on billions of edges
• you need supervised graph neural network training frameworks like PyTorch Geometric
• you require a permissive license since karateclub is GPL-3.0

Facets

library · maturity stable

machine-learning data-science nlp machine-learning data-science graph-processing python graph-embedding community-detection networkx node2vec deepwalk graph2vec louvain unsupervised-learning graph-mining network-science algorithms

Member repositories

For agents

markdown · JSON · MCP: product_card(name="benedekrozemberczki/karateclub")