Web Retrieval-Aware Chunking (W-RAC) for Efficient and Cost-Effective Retrieval-Augmented Generation Systems
Summary
W-RAC introduces a cost-efficient chunking framework for web document processing in RAG systems that reduces LLM token usage by an order of magnitude through structured content representation and retrieval-aware grouping decisions. The method decouples text extraction from semantic chunk planning, achieving comparable or better retrieval performance than traditional chunking approaches while minimizing hallucination risks.
View Cached Full Text
Cached at: 04/20/26, 08:26 AM
Paper page - Web Retrieval-Aware Chunking (W-RAC) for Efficient and Cost-Effective Retrieval-Augmented Generation Systems
Source: https://huggingface.co/papers/2604.04936
Abstract
Web Retrieval-Aware Chunking (W-RAC) introduces a cost-efficient framework for web document processing that reduces LLM token usage and hallucination risks through structured content representation and retrieval-aware grouping decisions.
Retrieval-Augmented Generation (https://huggingface.co/papers?q=Retrieval-Augmented%20Generation) (RAG) systems critically depend on effective document chunking (https://huggingface.co/papers?q=document%20chunking) strategies to balance retrieval quality, latency, and operational cost. Traditional chunking approaches, such as fixed-size, rule-based, or fully agentic chunking, often suffer from high token consumption (https://huggingface.co/papers?q=token%20consumption), redundant text generation, limited scalability, and poor debuggability, especially for large-scale web content ingestion. In this paper, we propose Web Retrieval-Aware Chunking (W-RAC), a novel, cost-efficient chunking framework designed specifically for web-based documents (https://huggingface.co/papers?q=web-based%20documents). W-RAC decouples text extraction from semantic chunk planning by representing parsed web content as structured, ID-addressable units and leveraging large language models (https://huggingface.co/papers?q=large%20language%20models) (LLMs) only for retrieval-aware grouping (https://huggingface.co/papers?q=retrieval-aware%20grouping) decisions rather than text generation. This significantly reduces token usage, eliminates hallucination risks (https://huggingface.co/papers?q=hallucination%20risks), and improves system observability. Experimental analysis and architectural comparison demonstrate that W-RAC achieves comparable or better retrieval performance than traditional chunking approaches while reducing chunking-related LLM costs by an order of magnitude.
View arXiv page (https://arxiv.org/abs/2604.04936) View PDF (https://arxiv.org/pdf/2604.04936) Add to collection (https://huggingface.co/login?next=%2Fpapers%2F2604.04936)
Get this paper in your agent:
hf papers read 2604.04936
Don’t have the latest CLI? curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2604.04936 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2604.04936 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2604.04936 in a Space README.md to link it from this page.
Collections including this paper 1
Similar Articles
Adaptive Chunking: Optimizing Chunking-Method Selection for RAG
Introduces Adaptive Chunking, a framework using five intrinsic document metrics to select optimal chunking strategies for RAG, improving answer correctness from 62-64% to 72% and question resolution rate by over 30%.
LFRAG: Layout-oriented Fine-grained Retrieval-Augmented Generation on Multimodal Document Understanding
LFRAG proposes a layout-oriented fine-grained retrieval-augmented generation framework that moves from page-level to block-level retrieval in multimodal documents, achieving state-of-the-art performance and 73% token reduction on the new LFDocQA benchmark.
LightRAG: Simple and Fast Retrieval-Augmented Generation
The article introduces LightRAG, an open-source framework that enhances Retrieval-Augmented Generation by integrating graph structures for improved contextual awareness and efficient information retrieval.
Query-Adaptive Semantic Chunking for Retrieval-Augmented Generation: A Dynamic Strategy with Contextual Window Expansion
Proposes a query-adaptive semantic chunking method for retrieval-augmented generation that dynamically adjusts chunk boundaries using contextual window expansion to improve retrieval precision.
MM-BizRAG: Rethinking Multimodal Retrieval-Augmented Generation for General Purpose Enterprise Q&A
MM-BizRAG is a multimodal retrieval-augmented generation system for enterprise Q&A that uses document structure-aware splitting and layout-aware parsing to outperform vision-centric baselines by up to 32% on heterogeneous enterprise documents. The paper also introduces FastRAGEval, a cost-efficient LLM-based evaluation metric with stronger human alignment than RAGChecker.