Dynamic Coattention Networks For Question Answering

Caiming Xiong, Victor Zhong, Richard Socher

Nov 04, 2016 (modified: Feb 14, 2017) ICLR 2017 conference submission readers: everyone
  • Abstract: Several deep learning models have been proposed for question answering. How- ever, 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 pointer 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.
  • TL;DR: An end-to-end dynamic neural network model for question answering that achieves the state of the art and best leaderboard performance on the Stanford QA dataset.
  • Keywords: Natural language processing, Deep learning, Applications
  • Conflicts: salesforce.com, metamind.io