This repository serves as a curated archive of academic papers focused on dynamic graph representation learning and analysis, particularly within social networks and knowledge graphs. It targets researchers and practitioners in graph machine learning, providing a comprehensive overview of foundational and state-of-the-art techniques. The collection aims to facilitate understanding and advancement in modeling evolving network structures and their associated dynamics.

How It Works

The project is a living collection of research papers, categorized by sub-fields such as static graph representation, dynamic graph representation, heterogeneous graphs, and dynamic heterogeneous graphs. Each entry typically includes paper titles, authors, publication venues, dates, keywords, brief summaries, and links to the original papers or code. The repository is actively maintained and encourages community contributions through GitHub issues.

Quick Start & Requirements

This repository is a collection of research papers and does not have a direct installation or execution command. Accessing the content requires browsing the README and following the provided links to academic papers and datasets.

Highlighted Details

• Extensive categorization of papers covering static, dynamic, heterogeneous, and dynamic heterogeneous graph representation learning.
• Includes foundational works like node2vec, LINE, and GCN, alongside recent advancements in temporal graph networks and attention mechanisms.
• Provides links to numerous datasets relevant to dynamic graph analysis, such as GDELT, ICEWS, and SNAP temporal datasets.
• Features a dedicated section for Large Language Models applied to dynamic graphs, reflecting current research trends.

Maintenance & Community

The project is maintained by "Cantoria" and actively encourages community engagement through GitHub's watch, star, and fork features. Users are invited to submit recommended works, download links, and highlight key contributions via GitHub issues.

Licensing & Compatibility

The repository itself does not specify a license. The linked academic papers are subject to their respective publisher's copyright and licensing terms.

Limitations & Caveats

As a curated list of papers, the repository does not provide executable code or a unified framework. Users must individually access, download, and potentially implement the methods described in the linked research papers. The "Updating" tags indicate sections that are actively being expanded.