Dynamic Coattention Networks for Question Answering
Caiming Xiong, Victor Zhong, Richard Socher Salesforce Research Palo Alto, CA 94301, USA {cxiong, vzhong, Equal contribution
Abstract
Several deep learning models have been proposed for question answering. However, due to their single-pass nature, they have no way to recover from local maxima corresponding to incorrect answers. To address this problem, we introduce the Dynamic Coattention Network (DCN) for question answering. The DCN first fuses co-dependent representations of the question and the document in order to focus on relevant parts of both. Then a dynamic pointing decoder iterates over potential answer spans. This iterative procedure enables the model to recover from initial local maxima corresponding to incorrect answers. On the Stanford question answering dataset, a single DCN model improves the previous state of the art from 71.0% F1 to 75.9%, while a DCN ensemble obtains 80.4% F1.
中文速览
机器阅读理解任务要求模型从一段文字中找出问题的答案，而以往的深度学习方法只做"单次预测"，一旦锁定了错误的答案片段就无法自我纠正。为此，研究者提出了动态协同注意力网络（Dynamic Coattention Network，DCN），它先通过协同注意力机制让问题和文档互相"看懂"对方——既根据文档内容理解问题，也根据问题内容聚焦文档的相关部分——再用一个动态指针解码器反复迭代地调整答案的起止位置，从而跳出错误的局部最优解。在斯坦福问答数据集（SQuAD）上，单个DCN模型将当时最优F1分数从71.0%提升至75.9%，集成模型更达到80.4%。这项工作意义在于它证明了"让模型带着上下文多想几遍"的迭代解码策略能显著提升答案定位的准确性，为后续阅读理解研究树立了重要基线。
原文 arXiv:1611.01604;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1611.01604v4