This repository serves as a curated collection of resources on Retrieval-Augmented Generation (RAG) with Graphs, targeting researchers and practitioners in Natural Language Processing (NLP) and Knowledge Graphs. It aims to consolidate papers, tools, and data sources, categorizing them by applied graph domains to facilitate understanding and advancement in GraphRAG techniques.

How It Works

The repository itself is a comprehensive, categorized bibliography of research papers and code related to GraphRAG. It does not contain executable code for GraphRAG systems but rather lists papers that explore various methods, such as augmenting LLMs with knowledge graphs for question answering, improving retrieval with graph structures, and applying graph neural networks in RAG pipelines. The organization by graph domain (Knowledge Graph, Document Graph, Scientific Graph, etc.) highlights the diverse applications and approaches within the field.

Quick Start & Requirements

This repository is a curated list of resources and does not have a direct installation or execution command. The primary "code" mentioned is get_hot.py, which likely serves to identify highly cited papers. Links to papers (arXiv, conference proceedings) and associated code repositories are provided throughout the README.

Highlighted Details

• Extensive categorization of GraphRAG research across multiple graph domains including Knowledge Graphs, Document Graphs, Scientific Graphs, Social Graphs, and more.
• Includes papers focusing on specific applications like question answering, recommendation systems, summarization, and fraud detection.
• Highlights highly cited papers (marked with :fire:) and provides links to associated code and benchmarks for various GraphRAG techniques.
• Covers foundational and advanced RAG techniques, benchmarks, and evaluations, offering a broad overview of the field's landscape.

Maintenance & Community

The repository is maintained by a team of contributors from the University of Oregon and Michigan State University. Users are encouraged to open issues or pull requests for errors or missed papers.

Licensing & Compatibility

The repository itself does not specify a license. Individual papers and code repositories linked within the README will have their own respective licenses. Compatibility for commercial use or closed-source linking would depend on the licenses of the linked resources.

Limitations & Caveats

This repository is a bibliography and does not provide a unified GraphRAG framework or implementation. Users must refer to individual linked papers and code repositories for specific functionalities, requirements, and licensing. The list is intended to be updated, but its comprehensiveness and recency depend on community contributions.