Traditional Retrieval Augmented Generation (RAG) relies on semantic search to retrieve relevant information from unstructured text documents. However, additional details based on relationships between data entities, which may not be semantically connected in general but are connected in the document, can make the search and generation more accurate. Graph RAG explicitly extracts entities and relationships from the unstructured text and represents them as a structured knowledge graph. 

Target…