arXiv 2606.139052026-06-15cs.IRcs.CL

ADORE: 基于检索反馈的迭代查询扩展

ADORE: Iterative Query Expansion with Retrieval-Grounded Relevance Feedback

University of Waterloo（滑铁卢大学）；Mila – Quebec AI Institute（魁北克人工智能研究所）；University of Toronto（多伦多大学）；University of California, Berkeley（加州大学伯克利分校）

Amin Bigdeli, Negar Arabzadeh, Radin Hamidi Rad, Sajad Ebrahimi, Charles L. A. Clarke, Ebrahim Bagheri

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出ADORE框架，通过迭代生成伪段落、检索语料库并评估相关性，利用检索反馈指导查询扩展，显著提升检索性能。

摘要：

基于LLM的查询扩展通过为原始查询添加额外上下文来改进检索。然而，大多数方法仍然是生成驱动的，产生看似合理的伪文档或扩展，而不检查目标语料库的响应。这可能导致检索漂移、放大误导性词汇或遗漏区分相关与不相关文档的术语。我们认为，有效的扩展需要基于检索的反馈，而不仅仅是单次生成或未经验证的迭代。我们引入ADORE（适应、观察、相关性评估），一个迭代框架，将检索结果转化为下一次扩展的反馈。在每一轮中，LLM生成伪段落，检索器暴露语料库响应，相关性评估器根据原始查询评估检索到的文档。这些判断识别出需要强化、仍未被覆盖以及需要抑制的内容。在TREC Deep Learning、BEIR和BRIGHT上，ADORE consistently outperforms strong query expansion baselines with notable improvements across nearly all evaluation settings, improving average nDCG@10 by 24.5% over BM25 and 3.6% over the strongest prior query expansion method on BEIR, and by 122.9% over BM25 and 9.2% over the best query expansion baseline on BRIGHT. 我们的代码和数据已公开。

英文摘要：

LLM-based query expansion improves retrieval by enriching the original query with additional context. Yet most methods remain generation-driven, producing plausible pseudo-documents or expansions without checking how the target corpus responds. This can introduce retrieval drift, amplify misleading vocabulary, or miss terms that distinguish relevant from non-relevant documents. We argue that effective expansion requires retrieval-grounded feedback, not just single-pass generation or unverified iteration. We introduce ADORE (ADapt, Observe, Relevance Evaluate), an iterative framework that turns retrieval outcomes into feedback for the next expansion. At each round, an LLM generates pseudo-passages, a retriever exposes the corpus response, and a relevance assessor evaluates retrieved documents against the original query. These judgments identify what to reinforce, what remains undercovered, and what to suppress. Across TREC Deep Learning, BEIR, and BRIGHT, ADORE consistently outperforms strong query expansion baselines with notable improvements across nearly all evaluation settings, improving average nDCG@10 by 24.5% over BM25 and 3.6% over the strongest prior query expansion method on BEIR, and by 122.9% over BM25 and 9.2% over the best query expansion baseline on BRIGHT. Our code and data are publicly available.

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。