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

facebooktwittermailshare

Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning

Abstract: 

This paper presents a new method --- adversarial advantage actor-critic (Adversarial A2C), which significantly improves the efficiency of dialogue policy learning in task-completion dialogue systems. Inspired by generative adversarial networks (GAN), we train a discriminator to differentiate responses/actions generated by dialogue agents from responses/actions by experts. Then, we incorporate the discriminator as another critic into the advantage actor-critic (A2C) framework, to encourage the dialogue agent to explore state-action within the regions where the agent takes actions similar to those of the experts. Experimental results in a movie-ticket booking domain show that the proposed Adversarial A2C can accelerate policy exploration efficiently.

up
0 users have voted:

Paper Details

Authors:
Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, Yun-Nung Chen, Kam-Fai Wong
Submitted On:
22 April 2018 - 12:00pm
Short Link:
Type:
Poster
Event:
Presenter's Name:
Yun-Nung Chen
Paper Code:
HLT-P2.4
Document Year:
2018
Cite

Document Files

Poster for Advantage A2C Dialogue Policy Learning

(41 downloads)

Subscribe

[1] Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, Yun-Nung Chen, Kam-Fai Wong, "Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3134. Accessed: Jul. 17, 2018.
@article{3134-18,
url = {http://sigport.org/3134},
author = {Baolin Peng; Xiujun Li; Jianfeng Gao; Jingjing Liu; Yun-Nung Chen; Kam-Fai Wong },
publisher = {IEEE SigPort},
title = {Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning},
year = {2018} }
TY - EJOUR
T1 - Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning
AU - Baolin Peng; Xiujun Li; Jianfeng Gao; Jingjing Liu; Yun-Nung Chen; Kam-Fai Wong
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3134
ER -
Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, Yun-Nung Chen, Kam-Fai Wong. (2018). Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning. IEEE SigPort. http://sigport.org/3134
Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, Yun-Nung Chen, Kam-Fai Wong, 2018. Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning. Available at: http://sigport.org/3134.
Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, Yun-Nung Chen, Kam-Fai Wong. (2018). "Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning." Web.
1. Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, Yun-Nung Chen, Kam-Fai Wong. Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3134