QChunker: Learning Question-Aware Text Chunking for Domain RAG via Multi-Agent Debate

TL;DR

QChunker introduces a multi-agent debate framework to improve text chunking for RAG systems, enhancing semantic integrity and information richness through question-driven segmentation and a novel evaluation metric.

Contribution

The paper presents a new multi-agent debate approach for question-aware text chunking and introduces ChunkScore, a direct evaluation metric for chunk quality, improving RAG performance.

Findings

QChunker produces more coherent and information-rich text chunks.

ChunkScore effectively evaluates chunk quality directly.

Experimental results show improved RAG performance across domains.

Abstract

The effectiveness upper bound of retrieval-augmented generation (RAG) is fundamentally constrained by the semantic integrity and information granularity of text chunks in its knowledge base. To address these challenges, this paper proposes QChunker, which restructures the RAG paradigm from retrieval-augmentation to understanding-retrieval-augmentation. Firstly, QChunker models the text chunking as a composite task of text segmentation and knowledge completion to ensure the logical coherence and integrity of text chunks. Drawing inspiration from Hal Gregersen's "Questions Are the Answer" theory, we design a multi-agent debate framework comprising four specialized components: a question outline generator, text segmenter, integrity reviewer, and knowledge completer. This framework operates on the principle that questions serve as catalysts for profound insights. Through this pipeline, we successfully construct a high-quality dataset of 45K entries and transfer this capability to small language models. Additionally, to handle long evaluation chains and low efficiency in existing chunking evaluation methods, which overly rely on downstream QA tasks, we introduce a novel direct evaluation metric, ChunkScore. Both theoretical and experimental validations demonstrate that ChunkScore can directly and efficiently discriminate the quality of text chunks. Furthermore, during the text segmentation phase, we utilize document outlines for multi-path sampling to generate multiple candidate chunks and select the optimal solution employing ChunkScore. Extensive experimental results across four heterogeneous domains exhibit that QChunker effectively resolves aforementioned issues by providing RAG with more logically coherent and information-rich text chunks.