Awesome Network Embedding

Description

Awesome Network Embedding is a curated directory of research papers and implementations related to network embedding, also known as network representation learning, graph embedding, and knowledge embedding. It focuses on methods for learning vertex (node) representations from graphs and networks.

Key Details

• Type: Curated GitHub list / themed directory
• Domain: Network embedding, graph representation learning, knowledge embedding
• Primary Content: Research paper references with links to corresponding code implementations
• Format: Markdown list organized by method/algorithm name

Features

• Curated paper list

  • Collects network embedding and related representation learning papers.
  • Includes classical and recent methods in graph and network representation learning.

• Multiple naming conventions covered

  • Explicitly includes methods under terms such as:
    • Network embedding
    • Network representation learning
    • Graph embedding
    • Knowledge embedding

• Method-level entries with references and code

  • Each listed method typically includes:
    • Method name
    • Full paper title
    • Publication venue and year
    • Link to the paper (e.g., arXiv, conference proceedings)
    • Link(s) to implementation(s), usually open source

• Examples of listed methods (with implementations)

  • GraphGym
    • A platform for designing and evaluating Graph Neural Networks (GNNs), NeurIPS 2020.
    • Links to the NeurIPS paper.
    • Python implementation repository.
  • FEATHER
    • "Characteristic Functions on Graphs: Birds of a Feather, from Statistical Descriptors to Parametric Models", CIKM 2020.
    • Paper link (arXiv).
    • Python implementations, including a KarateClub integration.
  • HeGAN
    • "Adversarial Learning on Heterogeneous Information Networks", KDD 2019.
    • Paper link.
    • Python implementation repository.
  • NetMF
    • "Network Embedding as Matrix Factorization: Unifying DeepWalk, LINE, PTE, and Node2Vec", WSDM 2018.
    • Paper link (matrix factorization view of popular network embedding methods).

• Implementation-focused organization

  • Emphasis on methods that have publicly available implementations.
  • Multiple implementation links possible per method (e.g., different Python packages).

• Community-updatable

  • Open to contributions via commits/pull requests for adding or reorganizing papers and methods.
  • Maintainer has indicated plans to further classify and reorganize entries with clearer indexing.

Category & Tags

• Category: Themed directories
• Tags:
  • machine-learning
  • graph
  • awesome-lists

Pricing

• Not a commercial product or service.
• Access to the list is free via GitHub (subject to repository’s license and GitHub’s terms).