A Coarse to Fine Question Answering System based on Reinforcement Learning

Abstract

In this paper, we present a coarse to fine question answering (CFQA) system based on reinforcement learning which can efficiently processes documents with different lengths by choosing appropriate actions. The system is designed using an actor-critic based deep reinforcement learning model to achieve multi-step question answering. Compared to previous QA models targeting on datasets mainly containing either short or long documents, our multi-step coarse to fine model takes the merits from multiple system modules, which can handle both short and long documents. The system hence obtains a much better accuracy and faster trainings speed compared to the current state-of-the-art models. We test our model on four QA datasets, WIKEREADING, WIKIREADING LONG, CNN and SQuAD, and demonstrate 1.3$\%$-1.7$\%$ accuracy improvements with 1.5x-3.4x training speed-ups in comparison to the baselines using state-of-the-art models.

arXiv comment9 pages, original work published in AAAI 2019

References in corpus (12)

• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• ADADELTA: An Adaptive Learning Rate Method
• SQuAD: 100,000+ Questions for Machine Comprehension of Text
• The Goldilocks Principle: Reading Children's Books with Explicit Memory Representations
• Dynamic Coattention Networks For Question Answering
• Deep Learning for Answer Sentence Selection
• A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task
• Language Understanding for Text-based Games Using Deep Reinforcement Learning
• NewsQA: A Machine Comprehension Dataset
• Improving Information Extraction by Acquiring External Evidence with Reinforcement Learning
• Who did What: A Large-Scale Person-Centered Cloze Dataset
• Improving the Speed of Response of Learning Algorithms Using Multiple Models