MM-BizRAG：面向通用企业问答的多模态检索增强生成再思考
MM-BizRAG: Rethinking Multimodal Retrieval-Augmented Generation for General Purpose Enterprise Q&A
AI总结：
提出MM-BizRAG框架，通过文档结构感知分割和布局感知解析，结合统一LLM驱动的工件转换与推理时多模态组装，无需微调即可提升企业文档问答性能，在异构企业数据集和两个公开基准上超越基线最多32个百分点。
摘要：
近期多模态检索增强生成（MM-RAG）的进展倾向于最小化解析，依赖页面级图像来生成检索器嵌入和答案生成。虽然高效，但这种趋势往往忽略了对复杂企业文档中丰富结构化信息的显式处理，而是依赖预训练嵌入或视觉语言模型隐式捕获这种结构。在本工作中，我们采取更直接的方法：MM-BizRAG通过文档结构感知分割主动提取和表示文档结构，该分割根据文档方向动态路由文档至特定方向的摄取管道，对垂直结构文档（如报告）应用显式布局感知解析，对水平结构文档（如幻灯片）应用整体页面级表示。统一的LLM驱动的工件转换管道通过基于占位符的位置对齐保留自然阅读顺序，而推理时的多模态组装将检索表示与生成上下文解耦，无需任何微调即可生成更丰富、更基于事实的答案。通过在大型异构企业数据集和两个公开基准（SlideVQA和FinRAGBench-V）上的实验，MM-BizRAG一致地超越最先进的以视觉为中心的基线最多32个百分点，在报告式布局上尤其强劲。此外，我们引入了FastRAGEval，一种单次调用的LLM评判指标，用于细粒度生成召回，将RAGChecker的成本减半，同时实现更强的人类对齐。
英文摘要：
Recent advances in multimodal retrieval-augmented generation (MM-RAG) have shifted toward minimal parsing, relying on page-level images for producing retriever embeddings and for answer generation. While efficient, this trend often neglects explicit handling of the rich, structured information in complex enterprise documents, instead depending on pre-trained embeddings or vision-language models to implicitly capture such structure. In this work, we take a more direct approach: MM-BizRAG proactively extracts and represents document structure via a document structure-aware split that dynamically routes documents through orientation-specific ingestion pipelines, applying explicit layout-aware parsing for vertically structured documents (e.g., reports) and holistic page-level representations for horizontally structured documents (e.g., slide decks). A unified LLM-driven artifact transformation pipeline with placeholder-based positional alignment preserves natural reading order, while inference-time multimodal assembly decouples retrieval representations from generation context, enabling richer, more grounded answers without any finetuning requirement. Through experiments on a large, heterogeneous enterprise dataset and two public benchmarks (SlideVQA and FinRAGBench-V), MM-BizRAG consistently outperforms state-of-the-art vision-centric baselines by up to 32% points, with especially strong gains on report-style layouts. Furthermore, we introduce FastRAGEval, a single-call LLM Judge metric for fine-grained generative recall that halves RAGChecker's cost while achieving stronger human alignment.
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。
