---
title: GraphRAG Local To Global
emoji: 💬
colorFrom: yellow
colorTo: purple
sdk: gradio
sdk_version: 4.36.1
app_file: app.py
pinned: false
license: mit
short_description: 'Microsoft''s GraphRAG implementation for a medical  book '
---

Implementation of Retriever and Generation of Microsoft Graph RAG.
Select search strategy as `Local` or `Global` from the Additional Inputs.
Input data: Diseases and Its Causes (https://www.gutenberg.org/ebooks/15283) from Project Gutenberg
Input data was indexed using `graphrag` library from microsoft. The results were stored in Neo4j.

Retrieval is done from the entity, relationships and communities in the indexed knowledge graph on Neo4j.