Query-Centric Graph Retrieval Augmented Generation
This addresses the problem of inefficient knowledge retrieval for multi-hop reasoning in LLMs, offering a novel paradigm that is not incremental.
The paper tackles the granularity dilemma in graph-based retrieval-augmented generation (RAG) for LLMs by introducing QCG-RAG, a query-centric framework that enables controllable granularity indexing and multi-hop chunk retrieval, resulting in improved question answering accuracy on benchmarks like LiHuaWorld and MultiHop-RAG.
Graph-based retrieval-augmented generation (RAG) enriches large language models (LLMs) with external knowledge for long-context understanding and multi-hop reasoning, but existing methods face a granularity dilemma: fine-grained entity-level graphs incur high token costs and lose context, while coarse document-level graphs fail to capture nuanced relations. We introduce QCG-RAG, a query-centric graph RAG framework that enables query-granular indexing and multi-hop chunk retrieval. Our query-centric approach leverages Doc2Query and Doc2Query{-}{-} to construct query-centric graphs with controllable granularity, improving graph quality and interpretability. A tailored multi-hop retrieval mechanism then selects relevant chunks via the generated queries. Experiments on LiHuaWorld and MultiHop-RAG show that QCG-RAG consistently outperforms prior chunk-based and graph-based RAG methods in question answering accuracy, establishing a new paradigm for multi-hop reasoning.