Cross-Granularity Hypergraph Retrieval-Augmented Generation for Multi-hop Question Answering

2025-08-15

AI Summary

In multi-hop question answering, existing retrieval-augmented generation (RAG) methods struggle to jointly capture semantic similarity and structural relationships among distributed knowledge: conventional RAG neglects fine-grained structural dependencies between entities, while graph-based retrieval approaches underemphasize textual semantic modeling. To address this, we propose a hypergraph-driven, cross-granularity RAG framework. It constructs an entity hypergraph where fine-grained entities serve as nodes and coarse-grained paragraphs as hyperedges; leverages hypergraph diffusion to jointly encode semantic and structural relevance; and introduces a jointly optimized retrieval-augmentation module. This framework unifies multi-granularity knowledge representation, enhancing both retrieval accuracy and efficiency. Extensive experiments on multiple benchmark datasets demonstrate significant improvements over state-of-the-art methods—achieving substantial gains in QA performance and accelerating retrieval speed by 6×.

Technology Category

• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Natural Language Processing: Question Answering

Application Category

• 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
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Multi-hop question answering (MHQA) requires integrating knowledge scattered across multiple passages to derive the correct answer. Traditional retrieval-augmented generation (RAG) methods primarily focus on coarse-grained textual semantic similarity and ignore structural associations among dispersed knowledge, which limits their effectiveness in MHQA tasks. GraphRAG methods address this by leveraging knowledge graphs (KGs) to capture structural associations, but they tend to overly rely on structural information and fine-grained word- or phrase-level retrieval, resulting in an underutilization of textual semantics. In this paper, we propose a novel RAG approach called HGRAG for MHQA that achieves cross-granularity integration of structural and semantic information via hypergraphs. Structurally, we construct an entity hypergraph where fine-grained entities serve as nodes and coarse-grained passages as hyperedges, and establish knowledge association through shared entities. Semantically, we design a hypergraph retrieval method that integrates fine-grained entity similarity and coarse-grained passage similarity via hypergraph diffusion. Finally, we employ a retrieval enhancement module, which further refines the retrieved results both semantically and structurally, to obtain the most relevant passages as context for answer generation with the LLM. Experimental results on benchmark datasets demonstrate that our approach outperforms state-of-the-art methods in QA performance, and achieves a 6$ imes$ speedup in retrieval efficiency.

Problem

• Integrates scattered knowledge for multi-hop question answering
• Balances structural and semantic retrieval in graph-based methods
• Improves retrieval efficiency and accuracy via hypergraph diffusion

Innovation

• Hypergraph integrates structural and semantic information
• Entity hypergraph links nodes and hyperedges
• Hypergraph diffusion enhances retrieval efficiency