InfoGain-RAG：通过基于文档信息增益的重排与过滤提升检索增强生成
InfoGain-RAG: Boosting Retrieval-Augmented Generation via Document Information Gain-based Reranking and Filtering
AI总结：
本文提出文档信息增益（DIG）指标量化检索文档对答案生成的贡献，并构建InfoGain-RAG框架训练专门重排器以过滤无关文档，在多个基准上显著优于现有RAG方法。
摘要：
检索增强生成（RAG）已成为解决大语言模型（LLM）关键局限性（如幻觉、知识过时和缺乏参考）的一种极具前景的方法。然而，当前的RAG框架通常难以识别检索到的文档是否对答案生成有实质性贡献。这一缺陷导致难以过滤掉不相关甚至具有误导性的内容，从而显著影响最终性能。在本文中，我们提出了文档信息增益（DIG），这是一种新颖的指标，旨在量化检索文档对正确答案生成的贡献。DIG通过计算LLM在有文档增强和无文档增强时生成置信度的差异来衡量文档的价值。此外，我们引入了InfoGain-RAG框架，该框架利用DIG分数训练专门的重新排序器，从精确区分和准确排序的角度对每个检索到的文档进行优先级排序。该方法能有效过滤无关文档并选择最有价值的文档，以生成更好的答案。在多个模型和基准上的广泛实验表明，InfoGain-RAG在单一和多重检索器范式下均显著优于现有方法。具体而言，在NaturalQA上，与朴素RAG、自反思RAG和现代基于排序的RAG相比，其精确匹配准确率分别提升了17.9%、4.5%和12.5%，在所有数据集上对高级专有模型GPT-4o平均提升达15.3%。这些结果证明了InfoGain-RAG的可行性，可为多种应用中的RAG提供可靠解决方案。
英文摘要：
Retrieval-Augmented Generation (RAG) has emerged as a promising approach to address key limitations of Large Language Models (LLMs), such as hallucination, outdated knowledge, and lacking reference. However, current RAG frameworks often struggle with identifying whether retrieved documents meaningfully contribute to answer generation. This shortcoming makes it difficult to filter out irrelevant or even misleading content, which notably impacts the final performance. In this paper, we propose Document Information Gain (DIG), a novel metric designed to quantify the contribution of retrieved documents to correct answer generation. DIG measures a document's value by computing the difference of LLM's generation confidence with and without the document augmented. Further, we introduce InfoGain-RAG, a framework that leverages DIG scores to train a specialized reranker, which prioritizes each retrieved document from exact distinguishing and accurate sorting perspectives. This approach can effectively filter out irrelevant documents and select the most valuable ones for better answer generation. Extensive experiments across various models and benchmarks demonstrate that InfoGain-RAG can significantly outperform existing approaches, on both single and multiple retrievers paradigm. Specifically on NaturalQA, it achieves the improvements of 17.9%, 4.5%, 12.5% in exact match accuracy against naive RAG, self-reflective RAG and modern ranking-based RAG respectively, and even an average of 15.3% increment on advanced proprietary model GPT-4o across all datasets. These results demonstrate the feasibility of InfoGain-RAG as it can offer a reliable solution for RAG in multiple applications.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
