This paper proposes a Markov Decision Process and reinforcement learning based approach for domain selection in a multi-domain Spoken Dialogue System built on a distributed architecture. In the proposed framework, the domain selection problem is treated as sequential planning instead of classification, such that confirmation and clarification interaction mechanisms are supported. In addition, it is shown that by using a model parameter tying trick, the extensibility of the system can be preserved, where dialogue components in new domains can be easily plugged in, without re-training the domain selection policy. The experimental results based on human subjects suggest that the proposed model marginally outperforms a non-trivial baseline.
|Title of host publication||Proceedings of EMNLP 2014|
|Subtitle of host publication||Conference on Empirical Methods in Natural Language Processing|
|Publisher||Association for Computational Linguistics|
|Publication status||Accepted/In press - 2014|
|Event||2014 Conference on Empirical Methods in Natural Language Processing - Doha, Qatar|
Duration: 25 Oct 2014 → 29 Oct 2014
|Conference||2014 Conference on Empirical Methods in Natural Language Processing|
|Abbreviated title||EMNLP 2014|
|Period||25/10/14 → 29/10/14|
Wang, Z., Chen, H., Wang, G., Tian, H., Wu, H., & Wang, H. (Accepted/In press). Policy Learning for Domain Selection in an Extensible Multi-domain Spoken Dialogue System. In Proceedings of EMNLP 2014: Conference on Empirical Methods in Natural Language Processing Association for Computational Linguistics.