Graph RAG, or Retrieval-Augmented Generation using knowledge graphs, has seen a significant increase in interest, with a 25% spike in Google searches in July 2024. This surge coincides with Microsoft’s announcement of their GraphRAG application on GitHub. Graph RAG combines the strengths of knowledge graphs, which store factual data, with Large Language Models (LLMs) to enhance the accuracy and context of responses. This technology has applications in customer service, drug discovery, regulatory reporting, HR, legal research, and wealth management. The interest in Graph RAG reflects its potential to solve complex, real-world problems. For instance, LinkedIn implemented Graph RAG, reducing their mean resolution time for customer service issues from 40 hours to 15 hours. The technology’s popularity is also evident from corporate moves like Samsung’s acquisition of RDFox and the merger of Ontotext and Semantic Web to form Graphwise, aiming to advance Graph RAG technology.
Source: towardsdatascience.com
