论文标题
在对话搜索中，将浓密的检索器免于快捷方式依赖性
Saving Dense Retriever from Shortcut Dependency in Conversational Search
论文作者
论文摘要
会话搜索（CS）需要对会话输入的整体理解以检索相关段落。在本文中，我们演示了CS中的检索快捷方式，该快捷方式使模型可以检索仅依靠部分历史的段落，同时忽略了最新问题。通过深入的分析，我们首先表明，训练有素的密集犬的天真密集的探索者大大利用了捷径，因此当被要求回答与历史无关的问题时，效果不佳。为了建立针对快捷依赖性的更强大的模型，我们探索了各种硬性销售策略。实验结果表明，基于模型的硬质量的训练有效地减轻了对快捷方式的依赖性，从而显着改善了最近CS基准的致密检索器。尤其是，我们的猎犬在QRECC上的召回@10中以11.0优于先前的最新模型。
Conversational search (CS) needs a holistic understanding of conversational inputs to retrieve relevant passages. In this paper, we demonstrate the existence of a retrieval shortcut in CS, which causes models to retrieve passages solely relying on partial history while disregarding the latest question. With in-depth analysis, we first show that naively trained dense retrievers heavily exploit the shortcut and hence perform poorly when asked to answer history-independent questions. To build more robust models against shortcut dependency, we explore various hard negative mining strategies. Experimental results show that training with the model-based hard negatives effectively mitigates the dependency on the shortcut, significantly improving dense retrievers on recent CS benchmarks. In particular, our retriever outperforms the previous state-of-the-art model by 11.0 in Recall@10 on QReCC.
