LightRAG: Simple and Fast Retrieval-Augmented Generation

LightRAG: простой и быстрый подход к генерации с дополнением извлечённой информацией
Lianghao Xia, Yanhua Yu, Zirui Guo, Tu Ao, Chao Huang
2025-01-01

LightRAGdual-level retrievalgraph-based retrievalincremental update algorithmretrieval-augmented generation
Retrieval-Augmented Generation (RAG) systems enhance large language models (LLMs) by integrating external knowledge sources, enabling more accurate and contextually relevant responses tailored to user needs.However, existing RAG systems have significant limitations, including reliance on flat data representations and inadequate contextual awareness, which can lead to fragmented answers that fail to capture complex interdependencies.To address these challenges, we propose Ligh-tRAG, a novel framework that incorporates graph structures into text indexing and retrieval processes.This innovative approach employs a dual-level retrieval system that enhances comprehensive information retrieval from both lowand high-level knowledge discovery.Additionally, the integration of graph structures with vector representations facilitates efficient retrieval of related entities and their relationships, significantly improving response times while maintaining contextual relevance.This capability is further enhanced by an incremental update algorithm that ensures the timely integration of new data, allowing the system to remain effective and responsive in rapidly changing data environments.Extensive experimental validation demonstrates considerable improvements in retrieval accuracy and efficiency compared to existing approaches.
1
An incremental update algorithm integrates new data promptly, maintaining responsiveness in rapidly changing knowledge environments.
2
Combining graph structures with vector representations enables efficient retrieval of related entities and relationships while preserving contextual relevance.
3
Experiments show considerable improvements in retrieval accuracy and efficiency compared with existing approaches.
4
Its dual-level retrieval mechanism supports comprehensive low-level and high-level knowledge discovery.
5
LightRAG introduces graph structures into text indexing and retrieval to address flat representations and weak contextual awareness in existing RAG systems.

LightRAG retrieval-augmented generation framework

Context-aware, efficient graph-based retrieval and incremental updating for accurate retrieval-augmented generation

Publication Details
Publication Date
2025-01-01
Journal
Publisher
ISSN
Cited by
828
Access Type
Author Information
Authors
Lianghao Xia
Yanhua Yu
Zirui Guo
Tu Ao
Chao Huang
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%