用信息片段生成回答，让回复更准确、可溯源且信息全。
GINGER: Grounded Information Nugget-Based Generation of Responses
- 将文档拆成最小信息单元，分步处理提升生成质量
- 在TREC RAG'24上表现超越现有方法，达成最新最优
- 适合需要高准确性和可解释性的问答系统使用
检索增强生成（RAG）面临事实准确性、来源归属和回复完整性挑战。为此，我们提出一种基于信息片段的模块化生成流程，该流程以从检索文档中提取的最小原子信息单元为基础。多阶段流程包括片段检测、聚类、排序、顶级聚类摘要与流畅性优化。该方法确保回复基于具体事实，支持来源追溯，并在长度限制下最大化信息包含量。在TREC RAG'24数据集上，通过AutoNuggetizer框架的广泛实验表明，GINGER在该基准上达到最先进的性能。
原文摘要 · Abstract (English)
Retrieval-augmented generation (RAG) faces challenges related to factual correctness, source attribution, and response completeness. To address them, we propose a modular pipeline for grounded response generation that operates on information nuggets-minimal, atomic units of relevant information extracted from retrieved documents. The multistage pipeline encompasses nugget detection, clustering, ranking, top cluster summarization, and fluency enhancement. It guarantees grounding in specific facts, facilitates source attribution, and ensures maximum information inclusion within length constraints. Extensive experiments on the TREC RAG'24 dataset evaluated with the AutoNuggetizer framework demonstrate that GINGER achieves state-of-the-art performance on this benchmark.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。