SkillQR improves query rewriting for better search results
SkillQR: Agentic Skill Discovery for Reason-Intensive Query Rewriting
Information Retrieval
Summary
Dense retrievers help find relevant information by matching meanings between a question and documents, but long or complicated questions can confuse them. The authors found that rewriting questions to fit the target search area works best when customized to that area. They created SkillQR, which figures out the best way to rewrite questions depending on where the search is done, without needing extra training data. SkillQR matches or beats some trained models while running faster than methods rewriting multiple times.
What this means in practice
- •For search engine engineers: Enhance search accuracy by dynamically choosing query rewrites tailored to specific document collections to improve retrieval relevance.
- •For medical data retrieval teams: Improve clinical information lookup by adapting query reformulations to specialized medical text collections, reducing irrelevant results.
Authors
Kenan Jiang, Meixiu Long, Zihao Zhang, Ran Xu, Eugene Agichtein, Fei Liu
Abstract
Dense retrievers provide grounded evidence that mitigates hallucinations in outputs from Large Language Models (LLMs). Real-world queries are often verbose, ambiguously framed, or require multi-step reasoning, introducing irrelevant details that weaken semantic matching for dense retrievers. Query rewriting is a widely adopted pre-retrieval technique that leverages LLMs to reformulate queries and bridge the semantic gap between such queries and the retrieval target. Existing rewriting approaches either rely on expensive proprietary models, substantial training data, or iterative retrieval and rewriting, while largely applying a uniform rewriting strategy regardless of the retrieval corpus. We observe that effective query rewriting instructions are not universal: the optimal strategy depends on the retrieval target, rather than the model alone. We introduce SkillQR, a query rewriting framework that automatically discovers reasoning strategy sets from the retrieval target and dynamically routes a query to the appropriate rewriting strategy at inference time. Without relying on training data, SkillQR achieves competitive performance compared to fine-tuned rewriting models on the BRIGHT and R2MED benchmarks, while also providing lower inference latency than iterative rewriting approaches.