POLICY COMMITTEE FOR ADAPTATION IN MULTI-DOMAIN SPOKEN DIALOGUE SYSTEMS

被引:0
作者
Gasic, M. [1 ]
Mrksic, N. [1 ]
Su, P-H. [1 ]
Vandyke, D. [1 ]
Wen, T-H. [1 ]
Young, S. [1 ]
机构
[1] Univ Cambridge, Dept Engn, Trumpington St, Cambridge CB2 1PZ, England
来源
2015 IEEE WORKSHOP ON AUTOMATIC SPEECH RECOGNITION AND UNDERSTANDING (ASRU) | 2015年
基金
英国工程与自然科学研究理事会;
关键词
Bayesian committee machines; Gaussian processes; reinforcement learning;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Moving from limited-domain dialogue systems to open domain dialogue systems raises a number of challenges. One of them is the ability of the system to utilise small amounts of data from disparate domains to build a dialogue manager policy. Previous work has focused on using data from different domains to adapt a generic policy to a specific domain. Inspired by Bayesian committee machines, this paper proposes the use of a committee of dialogue policies. The results show that such a model is particularly beneficial for adaptation in multi-domain dialogue systems. The use of this model significantly improves performance compared to a single policy baseline, as confirmed by the performed real-user trial. This is the first time a dialogue policy has been trained on multiple domains on-line in interaction with real users.
引用
收藏
页码:806 / 812
页数:7
相关论文
共 21 条
[21]   POMDP-Based Statistical Spoken Dialog Systems: A Review [J].
Young, Steve ;
Gasic, Milica ;
Thomson, Blaise ;
Williams, Jason D. .
PROCEEDINGS OF THE IEEE, 2013, 101 (05) :1160-1179