A user simulator architecture for socially-aware conversational agents

Alankar Jain, Florian Pecune, Yoichi Matsuyama, Justine Cassell

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Over the last two decades, Reinforcement Learning (RL) has emerged the method of choice for data-driven dialog management. However, one of the limitations of RL methods for the optimization of dialog managers in the context of virtual conversational agents, is that they require a large amount of data, which is often unavailable, particularly when the dialog deals with complex discourse phenomena. User simulators help address this problem by generating synthetic data to train RL agents in an online fashion. In this work, we extend user simulators to the case of socially-aware conversational agents, that combine task and social functions. We propose a novel architecture that takes into consideration the user’s conversational goals and generates both task and social behaviour. Our proposed architecture is general enough to be useful for training socially-aware conversational agents in any domain. As a proof of concept, we construct a user simulator for training a conversational recommendation agent and provide evidence towards the effectiveness of the approach.

Original languageEnglish
Title of host publicationProceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018
PublisherAssociation for Computing Machinery, Inc
Pages133-140
Number of pages8
ISBN (Electronic)9781450360135
DOIs
Publication statusPublished - 2018 Nov 5
Externally publishedYes
Event18th ACM International Conference on Intelligent Virtual Agents, IVA 2018 - Sydney, Australia
Duration: 2018 Nov 52018 Nov 8

Publication series

NameProceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018

Conference

Conference18th ACM International Conference on Intelligent Virtual Agents, IVA 2018
CountryAustralia
CitySydney
Period18/11/518/11/8

Fingerprint

Simulators
Reinforcement learning
Managers

Keywords

  • Conversational agent
  • Dialog system
  • Rapport
  • User simulator

ASJC Scopus subject areas

  • Artificial Intelligence
  • Human-Computer Interaction

Cite this

Jain, A., Pecune, F., Matsuyama, Y., & Cassell, J. (2018). A user simulator architecture for socially-aware conversational agents. In Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018 (pp. 133-140). (Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018). Association for Computing Machinery, Inc. https://doi.org/10.1145/3267851.3267916

A user simulator architecture for socially-aware conversational agents. / Jain, Alankar; Pecune, Florian; Matsuyama, Yoichi; Cassell, Justine.

Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018. Association for Computing Machinery, Inc, 2018. p. 133-140 (Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jain, A, Pecune, F, Matsuyama, Y & Cassell, J 2018, A user simulator architecture for socially-aware conversational agents. in Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018. Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018, Association for Computing Machinery, Inc, pp. 133-140, 18th ACM International Conference on Intelligent Virtual Agents, IVA 2018, Sydney, Australia, 18/11/5. https://doi.org/10.1145/3267851.3267916
Jain A, Pecune F, Matsuyama Y, Cassell J. A user simulator architecture for socially-aware conversational agents. In Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018. Association for Computing Machinery, Inc. 2018. p. 133-140. (Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018). https://doi.org/10.1145/3267851.3267916
Jain, Alankar ; Pecune, Florian ; Matsuyama, Yoichi ; Cassell, Justine. / A user simulator architecture for socially-aware conversational agents. Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018. Association for Computing Machinery, Inc, 2018. pp. 133-140 (Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018).
@inproceedings{d7dc6859cad6440b910d7c963d09cf2e,
title = "A user simulator architecture for socially-aware conversational agents",
abstract = "Over the last two decades, Reinforcement Learning (RL) has emerged the method of choice for data-driven dialog management. However, one of the limitations of RL methods for the optimization of dialog managers in the context of virtual conversational agents, is that they require a large amount of data, which is often unavailable, particularly when the dialog deals with complex discourse phenomena. User simulators help address this problem by generating synthetic data to train RL agents in an online fashion. In this work, we extend user simulators to the case of socially-aware conversational agents, that combine task and social functions. We propose a novel architecture that takes into consideration the user’s conversational goals and generates both task and social behaviour. Our proposed architecture is general enough to be useful for training socially-aware conversational agents in any domain. As a proof of concept, we construct a user simulator for training a conversational recommendation agent and provide evidence towards the effectiveness of the approach.",
keywords = "Conversational agent, Dialog system, Rapport, User simulator",
author = "Alankar Jain and Florian Pecune and Yoichi Matsuyama and Justine Cassell",
year = "2018",
month = "11",
day = "5",
doi = "10.1145/3267851.3267916",
language = "English",
series = "Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018",
publisher = "Association for Computing Machinery, Inc",
pages = "133--140",
booktitle = "Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018",

}

TY - GEN

T1 - A user simulator architecture for socially-aware conversational agents

AU - Jain, Alankar

AU - Pecune, Florian

AU - Matsuyama, Yoichi

AU - Cassell, Justine

PY - 2018/11/5

Y1 - 2018/11/5

N2 - Over the last two decades, Reinforcement Learning (RL) has emerged the method of choice for data-driven dialog management. However, one of the limitations of RL methods for the optimization of dialog managers in the context of virtual conversational agents, is that they require a large amount of data, which is often unavailable, particularly when the dialog deals with complex discourse phenomena. User simulators help address this problem by generating synthetic data to train RL agents in an online fashion. In this work, we extend user simulators to the case of socially-aware conversational agents, that combine task and social functions. We propose a novel architecture that takes into consideration the user’s conversational goals and generates both task and social behaviour. Our proposed architecture is general enough to be useful for training socially-aware conversational agents in any domain. As a proof of concept, we construct a user simulator for training a conversational recommendation agent and provide evidence towards the effectiveness of the approach.

AB - Over the last two decades, Reinforcement Learning (RL) has emerged the method of choice for data-driven dialog management. However, one of the limitations of RL methods for the optimization of dialog managers in the context of virtual conversational agents, is that they require a large amount of data, which is often unavailable, particularly when the dialog deals with complex discourse phenomena. User simulators help address this problem by generating synthetic data to train RL agents in an online fashion. In this work, we extend user simulators to the case of socially-aware conversational agents, that combine task and social functions. We propose a novel architecture that takes into consideration the user’s conversational goals and generates both task and social behaviour. Our proposed architecture is general enough to be useful for training socially-aware conversational agents in any domain. As a proof of concept, we construct a user simulator for training a conversational recommendation agent and provide evidence towards the effectiveness of the approach.

KW - Conversational agent

KW - Dialog system

KW - Rapport

KW - User simulator

UR - http://www.scopus.com/inward/record.url?scp=85058453593&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85058453593&partnerID=8YFLogxK

U2 - 10.1145/3267851.3267916

DO - 10.1145/3267851.3267916

M3 - Conference contribution

AN - SCOPUS:85058453593

T3 - Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018

SP - 133

EP - 140

BT - Proceedings of the 18th International Conference on Intelligent Virtual Agents, IVA 2018

PB - Association for Computing Machinery, Inc

ER -