DeepYardDeepYard
K

Knowledge Graph RAG with Citations

RAG pipeline using knowledge graphs for structured retrieval with source attribution

Open SourceFree

About

A RAG workflow that builds a knowledge graph from documents, enabling structured retrieval that preserves entity relationships and provides proper citations. Unlike vector-only RAG, knowledge graph RAG can answer multi-hop questions by traversing entity relationships, and every claim in the output links back to a specific source document and passage. Part of the awesome-llm-apps collection.

Details

Patternpipeline
StepsExtract entities and relationships from documents → Build knowledge graph with source tracking → Query graph with multi-hop traversal → Generate cited response from graph context
Use CasesResearch paper analysis, Compliance document Q&A, Multi-hop reasoning

Tags

ragknowledge-graphcitationsworkflow