Skip to main navigation Skip to search Skip to main content

Capsule Networks for Chinese Opinion Questions Machine Reading Comprehension

  • Longxiang Ding
  • , Zhoujun Li*
  • , Boyang Wang
  • , Yueying He
  • *Corresponding author for this work
  • Beihang University
  • National Computer Network Emergency Response Technical Team/Coordination Center of China

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In recent years, machine reading comprehension is becoming a more and more popular research topic. Promising results were obtained when the machine reading comprehension task had only two inputs, context and query. In this paper, we propose a capsule networks based model for Chinese opinion machine reading comprehension task which has three inputs: context, query and alternatives. First, we use a bi-directional LSTM to encode the three inputs. Second, model the complex interactions between context and query with a multiway attention layer. In addition to the attention mechanism used in BiDAF, the other two attention functions are designed to match the relationship between inputs. Finally, we present a capsule networks layer to route the right alternative. Specifically, we use two strategies to improve the dynamic routing process to filter noisy capsules, which may contain useless information such as stop words. Our single model achieves competitive results compared to the baseline methods on a Chinese dataset and obtains a significant improvement of 2.45% accuracy.

Original languageEnglish
Title of host publicationChinese Computational Linguistics - 18th China National Conference, CCL 2019, Proceedings
EditorsMaosong Sun, Yang Liu, Zhiyuan Liu, Xuanjing Huang, Heng Ji
PublisherSpringer
Pages521-532
Number of pages12
ISBN (Print)9783030323806
DOIs
StatePublished - 2019
Event18th China National Conference on Computational Linguistics, CCL 2019 - Kunming, China
Duration: 18 Oct 201920 Oct 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11856 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference18th China National Conference on Computational Linguistics, CCL 2019
Country/TerritoryChina
CityKunming
Period18/10/1920/10/19

Keywords

  • Capsule networks
  • Machine reading comprehension
  • Multiway attention

Fingerprint

Dive into the research topics of 'Capsule Networks for Chinese Opinion Questions Machine Reading Comprehension'. Together they form a unique fingerprint.

Cite this