EraRAG: Efficient and Incremental Retrieval Augmented Generation for Growing Corpora

2025-06-25

AI Summary

Existing Graph-RAG approaches rely on static graph structures, necessitating full graph reconstruction upon ingestion of new documents—rendering them ill-suited for dynamic corpora. To address this, we propose a multilevel graph framework supporting efficient incremental updates. Our method introduces hyperplane-based locality-sensitive hashing (LSH) into graph index construction for hierarchical, localized corpus organization and insertion. We further design an incremental graph update mechanism and retrieval path optimization strategy that eliminate the need for retraining or global graph reconstruction. Evaluated on large-scale benchmarks, our approach achieves nearly 10× faster update speed, substantially reduces computational overhead and token consumption, and maintains superior retrieval accuracy. By jointly optimizing real-time adaptability, scalability, and retrieval quality, our framework establishes a novel paradigm for dynamic knowledge integration in RAG systems.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Graph-based Retrieval-Augmented Generation (Graph-RAG) enhances large language models (LLMs) by structuring retrieval over an external corpus. However, existing approaches typically assume a static corpus, requiring expensive full-graph reconstruction whenever new documents arrive, limiting their scalability in dynamic, evolving environments. To address these limitations, we introduce EraRAG, a novel multi-layered Graph-RAG framework that supports efficient and scalable dynamic updates. Our method leverages hyperplane-based Locality-Sensitive Hashing (LSH) to partition and organize the original corpus into hierarchical graph structures, enabling efficient and localized insertions of new data without disrupting the existing topology. The design eliminates the need for retraining or costly recomputation while preserving high retrieval accuracy and low latency. Experiments on large-scale benchmarks demonstrate that EraRag achieves up to an order of magnitude reduction in update time and token consumption compared to existing Graph-RAG systems, while providing superior accuracy performance. This work offers a practical path forward for RAG systems that must operate over continually growing corpora, bridging the gap between retrieval efficiency and adaptability. Our code and data are available at https://github.com/EverM0re/EraRAG-Official.

Problem

• Efficient dynamic updates for growing corpora in Graph-RAG
• Avoid costly full-graph reconstruction with new documents
• Maintain high retrieval accuracy with low update latency

Innovation

• Multi-layered Graph-RAG framework for dynamic updates
• Hyperplane-based LSH for hierarchical graph structures
• Efficient localized insertions without retraining or recomputation