让对话搜索更稳定，应对数据分布变化时效果提升30%。
RCEM: Robust Conversational Search EMbedder in Distributional Shift
- 用大模型重写查询后对齐对话，简化匹配任务
- 在分布外场景下性能最高提升30%，减少过拟合
- 无需重新训练即可对接原有索引，适合实际部署
我们提出RCEM，一种在分布外变化下仍具鲁棒性的对话搜索嵌入模型。与以往直接学习对话到段落匹配的方法不同，RCEM将带有特殊标记的对话对齐至大模型重写的查询，同时保持原始嵌入空间不变。该设计使重写后的查询自动映射到相关段落。结果表明，RCEM（1）通过将长段落匹配简化为短查询匹配，降低过拟合；（2）训练无需对话-段落相关性标签；（3）保留原嵌入空间，可直接用于由原始嵌入器构建的索引，无需重建。大量实验显示，RCEM在分布外场景下持续优于先前方法，性能最高提升30%。
原文摘要 · Abstract (English)
We propose RCEM, a Robust Conversational search EMbedder that is additionally equipped with LLM's query reformulation capability without losing base model's generalization. Unlike prior conversational dense retrieval approaches that learn direct conversation-to-passage matching, RCEM aligns conversations, prepended by special token, to LLM-rewritten queries, while preserving the original embedding space. The unchanged embedding space automatically maps the rewritten-query to the relevant passages. As a result, RCEM (1) reduces overfitting by simplifying the alignment task from long passages to shorter rewritten queries, (2) eliminates the need for conversation-to-passage relevance labels for training, and (3) maintains its original embedding space that allows conversational queries against indexes built by original embedder without rebuilding them. Extensive experiments show that RCEM consistently outperforms prior approaches, achieving up to 30% improvement under distributional shift.
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。