Practical GraphRAG - Making LLMs Smarter with Knowledge Graphs

Explore the innovative approach of GraphRAG in enhancing Large Language Models (LLMs) with Knowledge Graphs in this 53-minute Devoxx conference talk. Dive into the limitations of traditional LLMs and discover how Retrieval-Augmented Generation (RAG) can provide current, relevant information to improve generative tasks. Learn how knowledge graphs offer a high-fidelity representation of data and reality, making rich context available based on user queries. Understand the process of transforming text data into graph data structures, leveraging LLM's language skills to extract entities and relationships. Examine the correlation of extracted information with sources, community clustering, and navigation techniques for answering questions. Gain insights into Microsoft Research's GraphRAG approach and witness live demonstrations of indexing and search using Neo4j and LangChain.

Practical GraphRAG - Making LLMs smarter with Knowledge Graphs by Michael Hunger