用大模型重排文档，让长文本生成覆盖更多关键信息。
LANCER: LLM Reranking for Nugget Coverage
- 通过预测子问题来引导文档重排，提升信息覆盖
- 在多个数据集上显著提升信息覆盖率和α-nDCG
- 适合需要全面信息的自动报告生成场景
与短文本检索增强生成（如事实问答）不同，长文本RAG要求检索提供广泛的相关信息。自动报告生成即为此类场景：不仅需要相关性，还需详尽全面的回答。然而现有检索方法主要优化相关性排序，而非信息覆盖。为此，我们提出LANCER，一种基于大模型的子问题覆盖重排方法。该方法预测满足信息需求所需的子问题，判断哪些文档回答了这些子问题，并据此重排文档，使排序结果覆盖尽可能多的信息点。实验表明，LANCER在信息覆盖率指标上优于其他大模型重排方法，显著提升α-nDCG与信息覆盖度。进一步的最优分析显示，子问题生成是核心关键。
原文摘要 · Abstract (English)
Unlike short-form retrieval-augmented generation (RAG), such as factoid question answering, long-form RAG requires retrieval to provide documents covering a wide range of relevant information. Automated report generation exemplifies this setting: it requires not only relevant information but also a more elaborate response with comprehensive information. Yet, existing retrieval methods are primarily optimized for relevance ranking rather than information coverage. To address this limitation, we propose LANCER, an LLM-based reranking method for nugget coverage. LANCER predicts what sub-questions should be answered to satisfy an information need, predicts which documents answer these sub-questions, and reranks documents in order to provide a ranked list covering as many information nuggets as possible. Our empirical results show that LANCER enhances the quality of retrieval as measured by nugget coverage metrics, achieving higher $α$-nDCG and information coverage than other LLM-based reranking methods. Our oracle analysis further reveals that sub-question generation plays an essential role.
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。