ericblue/knowledge-graph-llm-links

A collection of interesting links, articles, research papers and projects related to knowledge graphs, GenAI and LLMs (large language models)

Knowledge Graphs + LLMs

Background

A collection of interesting links, articles, research papers and projects related to knowledge graphs, GenAI and LLMs (large language models)

I’ve been fascinated with the concept of “connected knowledge” and knowledge representation for quite some time. When blogging was still a major trend, I wrote extensively about mind mapping and it’s still a significant part of my personal and business workflow today.

In the mid-late 2000s, I developed a keen interest in personal knowledge management. Inspired by the work of Gordon Bell (https://en.wikipedia.org/wiki/Gordon_Bell), I created my own Personal Memex project using Semantic Mediawiki (RDF, SPARQL, etc.) to store and visualize knowledge and data (https://eric-blue.com/my-projects/personal-memex/).

Fast forward to the last 18-24 months, and I’ve immersed myself in the GenAI space and large language models (LLMs). A couple of months ago, I posted about the National Security Hackathon in San Francisco (May - 2024), where we leveraged graph databases and large language models for knowledge extraction and modeling. Since then, partly due to the influence of social media algorithms and general momentum in this space, I’ve noticed a surge in activity. Although I’ve been aware of graph databases for some time and have experimented with them over the years, my interest has been reignited given their potential when coupled with GPT.

With that in mind, I wanted to share a collection of interesting links, articles, research papers, and projects related to knowledge graphs, GenAI, and LLMs. This list is not exhaustive and reflects the activity I’ve noted since May. However, I plan to continually update it. If anyone has recommendations for interesting projects, research, or key players in this space, please let me know—I’d love to include them in this list.

Books

Knowledge Graph-Enhanced RAG

• https://www.manning.com/books/knowledge-graph-enhanced-rag

Frameworks

Graph Databases

• Neo4j
  • https://neo4j.com/
• ArangoDB
  • https://arangodb.com/
• FalkorDB
  • https://github.com/FalkorDB
  • objective is to create an outstanding Knowledge Graph specifically for Large Language Models (LLM) that boasts exceptionally low latency, ensuring swift delivery of information through our Graph Database, known as KG-RAG.

Microsoft GraphRag

• Github
  • https://github.com/microsoft/graphrag
• Microsoft
  • modular graph-based Retrieval-Augmented Generation (RAG) system
• Links
  • GraphRAG: Unlocking LLM discovery on narrative private data
    • https://www.microsoft.com/en-us/research/blog/graphrag-unlocking-llm-discovery-on-narrative-private-data/

LlamaIndex

• Property Graph Index
  • Introducing the Property Graph Index: A Powerful New Way to Build Knowledge Graphs with LLMs
    • https://www.llamaindex.ai/blog/introducing-the-property-graph-index-a-powerful-new-way-to-build-knowledge-graphs-with-llms
    • integrates with Neo4j
• Articles
  • Customizing property graph index in LlamaIndex
    • https://www.llamaindex.ai/blog/customizing-property-graph-index-in-llamaindex
    • Github
      • https://github.com/tomasonjo/blogs/blob/master/llm/llama_index_neo4j_custom_retriever.ipynb

LangGraph

• https://langchain-ai.github.io/langgraph/
• Building language agents as graphs
  • library for building stateful, multi-actor applications with LLMs, used to create agent and multi-agent workflows
• Articles
  • Build a Reliable RAG Agent using LangGraph
    • https://medium.com/the-ai-forum/build-a-reliable-rag-agent-using-langgraph-2694d55995cd
• LangGraph Cloud
  • https://langchain-ai.github.io/langgraph/cloud/
  • Article
    • https://cobusgreyling.medium.com/langchain-just-launched-langgraph-cloud-bf8f65e45a54
  • Video
    • https://www.youtube.com/watch?v=l4sMKF1dTDM

Neo4j GraphRAG

• Video
  • Demo LLM Knowledge Graph Construction with Gemini (short)
    • https://www.youtube.com/watch?v=PeYwFoYY3jI
    • Transforms unstructured documents, Youtube videos, Wikipedia pages to Knowledge Graph in 𝟑 𝐜𝐥𝐢𝐜𝐤𝐬.
• Articles
  • Neo4j Brings GraphRAG Capabilities for GenAI to Google Cloud
    • https://neo4j.com/blog/graphrag-genai-googlecloud/
• GenAI package for Python
  • https://neo4j.com/docs/neo4j-genai-python/current/

DataBricks

• Neo4j-Databricks Connector Delivers Deeper Insights, Faster GenAI Development
  • https://neo4j.com/blog/neo4j-databricks-connector/

Research Articles / Papers

Cluster of research abstracts on graphs and llms

• Nomic Graph Research
  • https://atlas.nomic.ai/data/ct/graph-research/map/52484ea8-e730-480a-990d-ab917df980eb

A Survey on Open Information Extraction from Rule-based Model to Large Language Model

• https://arxiv.org/pdf/2208.08690v6
• A new survey paper traces the evolution of Open Information Extraction. - Open Information Extraction (OpenIE)

GraphReader

• The Next Level of Long-form Document Understanding and Why Iteration Matters
  • GraphReader with a 4k context window consistently outperformed GPT-4 with a 128k context window on texts ranging from 16k to 256k words
• GraphReader: Building Graph-based Agent to Enhance Long-Context Abilities of Large Language Models
  • https://arxiv.org/html/2406.14550v1

A Survey of Large Language Models for Graphs

• https://arxiv.org/abs/2405.08011
• Combining GNNs (Graph Neural Networks) with LLMs (Large Language Models)

GNN-RAG: Graph Neural Retrieval for Large Language Model Reasoning

• https://arxiv.org/abs/2405.20139

XLLM

• New Generation of Large Language Models
  • https://mltechniques.com/2024/06/03/new-trends-in-llm-overview-with-focus-on-xllm/
  • Presentation
    • Vincent Granville, PhD - Chief AI Scientist
    • https://docs.google.com/presentation/d/15jlAz0pOmybTxAzywzXklBcL1DLvQy50/edit#slide=id.p1
  • xLLM has been using knowledge graphs since the very beginning: taxonomies, related concepts (related links found in any large repository such as Wikipedia or corporate, that you can extract to enrich your LLM)

ML Techniques

• Resources & Papers
  • https://mltechniques.com/resources/

Unifying Large Language Models and Knowledge Graphs: A Roadmap

• https://arxiv.org/abs/2306.08302

From LLM Graph

• Knowledge Graph Generation From Text
  • https://arxiv.org/abs/2211.10511
• Towards Foundation Models for Knowledge Graph Reasoning
  • https://arxiv.org/abs/2310.04562
• BertNet: Harvesting Knowledge Graphs with Arbitrary Relations from Pretrained Language Models
  • https://arxiv.org/abs/2206.14268
• Large Language Models on Graphs: A Comprehensive Survey
  • https://arxiv.org/abs/2312.02783
• A Hitchhiker’s Guide to Knowledge Galaxies
  • https://caminao.blog/knowledge-management-booklet/a-hitchhikers-guide-to-knowledge-galaxies/
• A curated list of papers and resources based on "Large Language Models on Graphs: A Comprehensive Survey"
  • https://github.com/PeterGriffinJin/Awesome-Language-Model-on-Graphs

Notable People - Sources of Knowledge

LinkedIn

• Chandler T. Wilson
  • https://www.linkedin.com/in/chandlertwilson/
  • AI, LLMs, OSINT, and alternative data for corporate strategy, private equity, and public affairs
• Pascal Biese
  • https://www.linkedin.com/in/pascalbiese/
  • LLM Watch - Daily LLM highlights
• Tomaz Bratanic
  • https://www.linkedin.com/in/tomaz-bratanic-a58891127/
  • Graph ML and GenAI research at Neo4j
• Cobus Greyling
  • https://www.linkedin.com/in/cobusgreyling/
  • LLMs, NLP, NLU, Chatbots, Voicebots, CCAI, Ambient Orchestration, Ubiquitous User Interfaces
• Ryan Siegler
  • https://www.linkedin.com/in/ryan-siegler-816207102/
  • GenAI | Vector DBs | Data Science | Emerging Technology Advocate
• Dan Selman
  • https://www.linkedin.com/in/dselman/
  • Distinguished Engineer @ DocuSign | CTO, Chief Architect

LLM Watch

• https://www.llmwatch.com/
• Newsletter
  • https://www.linkedin.com/newsletters/llm-watch-7106973553371099136/

DAIR.AI

• Top AI-ML papers of the week
• Democratizing Artificial Intelligence Research, Education, and Technologies
• https://www.linkedin.com/company/dair-ai/posts/

Interesting Apps

Connected Papers

• Explore connected papers in a visual graph
• https://www.connectedpapers.com/

Open Source

• ScrapeGraphAI
  • https://github.com/VinciGit00/Scrapegraph-ai
  • https://scrapegraphai.com/
  • web scraping python library that uses LLM and direct graph logic to create scraping pipelines for websites and local documents (XML, HTML, JSON, etc.).
• Lab-Concerto-Graph
  • Graph storage for Concerto Models
  • https://github.com/accordproject/lab-concerto-graph
  • Concerto
    • https://concerto.accordproject.org/
  • Article
    • Text to Knowledge Graph
      • https://blog.selman.org/2024/06/22/text-to-knowledge-graph/
• Movie - Knowledge graph demo
  • https://github.com/dselman/movie-graph
  • Knowledge graph in 100 lines of code
    • https://www.docusign.com/blog/developers/knowledge-graph-100-lines-code?
• Llama Parse
  • Knowledge Graph Agent with LlamaParse
    • https://github.com/run-llama/llama_parse/blob/main/examples/knowledge_graphs/kg_agent.ipynb
• Chat Graph
  • A generic web interface to chat with any graph created with Concerto Graph
  • https://github.com/dselman/chat-graph
  • Video
    • https://www.youtube.com/watch?v=GtbEiIce-S0
  • Knowledge Graphs: RAG is NOT all you need
    • https://blog.selman.org/2024/06/04/knowledge-graphs-rag-is-not-all-you-need/
• R2R - production ready RAG engine w/ knowledge graphs
  • https://github.com/SciPhi-AI/R2R
    • https://github.com/SciPhi-AI/R2R
• EKnowledge
  • https://github.com/chigwell/eknowledge
  • Generating Knowledge Graphs from Textual Inputs Using Diverse Language Models
    • https://medium.com/@chigwel/eknowledge-generating-knowledge-graphs-from-textual-inputs-using-diverse-language-models-dfd8bf0f7b38
• LLM Graph
  • https://github.com/dylanhogg/llmgraph
• Graph Notebook: easily query and visualize graphs
  • https://github.com/aws/graph-notebook
• PyG re-implementation of Neural Bellman-Ford Networks
  • https://github.com/KiddoZhu/NBFNet-PyG

Memary

• https://www.memarylabs.com/
• Memory for Agent Self-Improvement
  • Agents promote human-like reasoning and represent a significant leap towards AGI and understanding ourselves as humans. Memory is a key component of how we as humans approach tasks and should be given equal importance when building AI agents. Memary pioneers the use of action graphs to track agent resources so that your agent can continuously self-improve in production environments and become more reliable over time.
• Github
  • https://github.com/kingjulio8238/memary

YWorks

• Advanced library for graph visualization
  • Fraud detection demo
    • https://www.yworks.com/demos/showcase/frauddetection/