Transformer-based pre-trained models like BERT have achieved great progress on Semantic Sentence Matching. Meanwhile, dependency prior knowledge has also shown general benefits in multiple NLP tasks. However, how to efficiently integrate dependency prior structure into pre-trained models to better model complex semantic matching relations is still unsettled. In this paper, we propose the \textbf{D}ependency-Enhanced \textbf{A}daptive \textbf{F}usion \textbf{A}ttention (\textbf{DAFA}), which explicitly introduces dependency structure into pre-trained models and adaptively fuses it with semantic information. Specifically, \textbf{\emph{(i)}} DAFA first proposes a structure-sensitive paradigm to construct a dependency matrix for calibrating attention weights. It adopts an adaptive fusion module to integrate the obtained dependency information and the original semantic signals. Moreover, DAFA reconstructs the attention calculation flow and provides better interpretability. By applying it on BERT, our method achieves state-of-the-art or competitive performance on 10 public datasets, demonstrating the benefits of adaptively fusing dependency structure in semantic matching task.

翻译：基于Transformer的预训练模型（如BERT）已在语义句子匹配任务中取得显著进展。同时，依赖先验知识也在多项自然语言处理任务中展现出通用优势。然而，如何高效地将依赖先验结构融入预训练模型以更好建模复杂语义匹配关系仍尚未解决。本文提出**依赖增强自适应融合注意力机制（DAFA）**，该机制显式地将依赖结构引入预训练模型，并使其与语义信息自适应融合。具体而言：**（i）** DAFA首先提出一种结构敏感范式，构建用于校准注意力权重的依赖矩阵；采用自适应融合模块整合所获取的依赖信息与原始语义信号。此外，DAFA重构注意力计算流程并提供更好的可解释性。通过将该方法应用于BERT，我们在10个公开数据集上取得最优或具有竞争力的性能，验证了在语义匹配任务中自适应融合依赖结构的优势。