Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager

Abstract: 

Natural language understanding and dialogue policy learning are both essential in conversational systems that predict the
next system actions in response to a current user utterance. Conventional approaches aggregate separate models of natural language understanding (NLU) and system action prediction (SAP) as a pipeline that is sensitive to noisy outputs of error-prone NLU. To address the issues, we propose an end-to-end deep recurrent neural network with limited contextual dialogue memory by jointly training NLU and SAP on DSTC4 multi-domain human-human dialogues. Experiments show that our proposed model significantly outperforms the state-of-the-art pipeline models for both NLU and SAP, which indicates that our joint model is capable of mitigating the affects of noisy NLU outputs, and NLU model can be refined by error flows backpropagating from the extra supervised signals of system actions.

up
0 users have voted:

Paper Details

Authors:
Xuesong Yang, Yun-Nung Chen, Dilek Hakkani-Tur, Paul Crook, Xiujun Li, Jianfeng Gao, Li Deng
Submitted On:
10 March 2017 - 2:14pm
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Xuesong Yang
Paper Code:
HLT-L2.4
Document Year:
2017
Cite

Document Files

E2E_ICASSP17.pdf

(65 downloads)

Subscribe

[1] Xuesong Yang, Yun-Nung Chen, Dilek Hakkani-Tur, Paul Crook, Xiujun Li, Jianfeng Gao, Li Deng, "End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1736. Accessed: Oct. 19, 2017.
@article{1736-17,
url = {http://sigport.org/1736},
author = {Xuesong Yang; Yun-Nung Chen; Dilek Hakkani-Tur; Paul Crook; Xiujun Li; Jianfeng Gao; Li Deng },
publisher = {IEEE SigPort},
title = {End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager},
year = {2017} }
TY - EJOUR
T1 - End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager
AU - Xuesong Yang; Yun-Nung Chen; Dilek Hakkani-Tur; Paul Crook; Xiujun Li; Jianfeng Gao; Li Deng
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1736
ER -
Xuesong Yang, Yun-Nung Chen, Dilek Hakkani-Tur, Paul Crook, Xiujun Li, Jianfeng Gao, Li Deng. (2017). End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager. IEEE SigPort. http://sigport.org/1736
Xuesong Yang, Yun-Nung Chen, Dilek Hakkani-Tur, Paul Crook, Xiujun Li, Jianfeng Gao, Li Deng, 2017. End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager. Available at: http://sigport.org/1736.
Xuesong Yang, Yun-Nung Chen, Dilek Hakkani-Tur, Paul Crook, Xiujun Li, Jianfeng Gao, Li Deng. (2017). "End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager." Web.
1. Xuesong Yang, Yun-Nung Chen, Dilek Hakkani-Tur, Paul Crook, Xiujun Li, Jianfeng Gao, Li Deng. End-to-End Joint Learning of Natural Language Understanding and Dialogue Manager [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1736