phanein/deepwalk

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

2759 stars · 812 forks observed · 2026-08-28

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

DeepWalk is a Python command-line tool that learns low-dimensional vector representations of vertices in graphs using short random walks and skipgram training. It takes graphs as adjacency lists, edge lists, or Matlab adjacency matrices and outputs node embeddings suitable for downstream machine learning tasks like multi-label node classification.

Use cases

• generate node embeddings for a graph
• learn graph representations with random walks
• embed social network nodes for classification
• convert a graph into feature vectors for machine learning
• run deepwalk on an adjacency list or edge list
• evaluate graph embeddings on multi-label node classification

When to choose

• you need classic DeepWalk node embeddings for graph mining or research
• you want a simple CLI that turns graph files into embedding vectors
• you are reproducing the DeepWalk paper's results on datasets like BlogCatalog
• you need unsupervised representations for nodes in a social or information network

When to avoid

• you need modern graph neural network methods like node2vec, GCN, or GraphSAGE
• you want an actively developed library with recent features and support
• you need a Python API rather than a command-line tool
• your graphs are so large they require distributed or GPU-accelerated training

Facets

cli-tool · maturity maintenance

machine-learning cli data-science machine-learning graph-processing data-science python cli windows graph-embeddings network-embeddings random-walks node-classification word2vec skipgram social-networks unsupervised-learning algorithms linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="phanein/deepwalk")