根据查询复杂度动态选择重排策略，显著降低延迟。
Adaptive Re-Ranking
- 按查询复杂度智能路由，选择轻量或重型重排模型
- 平均延迟降低1.11至5.22倍，最高降低53倍
- 适合对响应速度敏感的搜索系统部署
现代信息检索系统通常采用“检索-重排”流程，即用计算开销大的交叉编码器对快速初始检索器的结果进行重排。然而，该方法对所有查询统一使用高成本重排模型，导致简单查询也产生大量冗余计算和延迟。本文提出自适应重排（Adaptive Re-Ranking），基于效用的标签框架实现成本感知路由，并通过模拟最优路由与训练基线路由验证了按查询路由的巨大潜力，但学习过程受限于稀疏标注。我们训练了一个路由分类器，结合三种策略：稀疏检索（BM25）、密集重排（MiniLM-L6-v2）和重型神经重排（BGE-v2-m3）。相比BGE，本方法在所有测试数据集上实现1.15至53倍的中位延迟降低，平均延迟降低1.11至5.22倍，nDCG@10在-17.5%到+4.0%之间，部分数据集表现仍具竞争力。结果表明，基于新效用函数的查询路由为多种检索系统提供了可扩展的降本减时方案。
原文摘要 · Abstract (English)
Modern Information Retrieval (IR) systems typically use a "retrieve-then-rerank" pipeline, where a computationally expensive, pre-determined cross-encoder re-ranks the top results from a fast initial retriever. While effective, this approach often applies heavy re-ranking models regardless of query complexity, resulting in high latency and wasted computational resources on simple queries. We propose Adaptive Re-Ranking, an utility-based labeling framework for cost-aware routing and present empirical evidence (via oracle analysis and a trained baseline router) that per-query routing offers large potential gains but is non-trivial to learn from limited supervision. We train a routing classifier with 3 strategies: sparse retrieval (BM25), dense re-ranking (MiniLM-L6-v2), and heavy neural re-ranking (BGE-v2-m3). Compared to BGE our method achieves 1.15-53x lower median latency and 1.11-5.22x lower mean latency across all datasets we have tested, while delivering -17.5% to +4.0% nDCG@10, which is competitive in some datasets. Our findings show that routing queries based on our novel utility function offers a scalable solution for reducing computational costs and latency in a variety of IR systems.
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。