三阶段检索系统提升多轮问答准确率，性能优于基线10.7%。
Caraman at SemEval-2026 Task 8: Three-Stage Multi-Turn Retrieval with Query Rewriting, Hybrid Search, and Cross-Encoder Reranking
- 分三步：重写问题、混合检索、交叉编码重排
- 测试集nDCG@5达0.531，排名第八，超基线10.7%
- 技术领域用确定性生成，通用领域用随机性生成效果更优
我们介绍参与SemEval-2026 Task 8（MTRAGEval）Task A（检索）的系统，覆盖四个英文领域。采用三阶段流程：(1) 使用LoRA微调的Qwen 2.5 7B模型进行查询重写，将上下文依赖的后续问题转为独立查询；(2) 通过倒数排序融合（Reciprocal Rank Fusion）结合BM25与稠密检索；(3) 使用BGE-reranker-v2-m3进行交叉编码重排。在官方测试集上，系统nDCG@5达到0.531，38个参赛系统中排名第8，比主办方基线高出10.7%。开发集对比显示，针对不同领域调整生成温度——技术类用确定性解码，通用类用可控随机性——带来稳定提升；而更复杂的策略如领域感知提示或多查询扩展反而降低性能。
原文摘要 · Abstract (English)
We describe our system for SemEval-2026 Task 8 (MTRAGEval), participating in Task A (Retrieval) across four English-language domains. Our approach employs a three-stage pipeline: (1) query rewriting via a LoRA-fine-tuned Qwen 2.5 7B model that transforms context-dependent follow-up questions into standalone queries, (2) hybrid BM25 and dense retrieval combined through Reciprocal Rank Fusion, and (3) cross-encoder reranking with BGE-reranker-v2-m3. On the official test set, the system achieves nDCG@5 of 0.531, ranking 8th out of 38 participating systems and 10.7% above the organizer baseline. Development comparisons reveal that domain-specific temperature tuning for query generation, where technical domains benefit from deterministic decoding and general domains from controlled randomness, provides consistent gains, while more complex strategies such as domain-aware prompting and multi-query expansion degrade performance.
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。