Skip to main navigation Skip to search Skip to main content

An end-to-end neural dialog state tracking for task-oriented dialogs

  • Kyungpook National University
  • NAVER Corporation
  • Kyung Hee University

Research output: Contribution to conferenceConference paperpeer-review

Abstract

Dialog state tracking in spoken dialog system is the task that tracks the flow of a dialog and grasps what a user wants from the utterance precisely. Since the dialog success is related to catching the want of the user, dialog state tracking is a necessary component for spoken dialog systems. This paper proposes a neural dialog state tracker with the attention mechanism for focusing on valuable words and the hierarchical softmax for efficient training of the tracker. In addition, the proposed tracker combines a natural language understanding module and a dialog state module in an end-to-end style. As a result, the error propagation within a dialog system is minimized. To prove the effectiveness of the proposed model, we do experiments on dialog state tracking in the human-human task-oriented dialogs. Our experimental results show that the proposed method outperforms both the neural tracker without the attention mechanism and that without the hierarchical softmax.

Original languageEnglish
Title of host publication2018 IEEE International Conference on Fuzzy Systems, FUZZ-IEEE 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509060207
DOIs
StatePublished - 2018.10.12
Event2018 IEEE International Conference on Fuzzy Systems, FUZZ-IEEE 2018 - Rio de Janeiro, Brazil
Duration: 2018.07.82018.07.13

Publication series

NameIEEE International Conference on Fuzzy Systems
Volume2018-July
ISSN (Print)1098-7584

Conference

Conference2018 IEEE International Conference on Fuzzy Systems, FUZZ-IEEE 2018
Country/TerritoryBrazil
CityRio de Janeiro
Period18.07.818.07.13

Fingerprint

Dive into the research topics of 'An end-to-end neural dialog state tracking for task-oriented dialogs'. Together they form a unique fingerprint.

Cite this