Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents

Simon Keizer; Markus Guhe; Heriberto Cuayáhuitl; Ioannis  Efstathiou; Klaus-Peter Engelbrecht; Mihai Dobre; Alexandra Lascarides; Oliver Lemon

Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents

Simon Keizer, Markus Guhe, Heriberto Cuayáhuitl, Ioannis Efstathiou, Klaus-Peter Engelbrecht, Mihai Dobre, Alexandra Lascarides, Oliver Lemon

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In this paper we present a comparative evaluation of various negotiation strategies within an online version of the game “Settlers of Catan”. The comparison is based on human subjects playing games against artificial game-playing agents (‘bots’) which implement different negotiation dialogue strategies, using a chat dialogue interface to negotiate trades. Our results suggest that a negotiation strategy that uses persuasion, as well as a strategy that is trained from data using Deep Reinforcement Learning, both lead to an improved win rate against humans, compared to previous rule-based and supervised learning baseline dialogue negotiators.

Original language	English
Title of host publication	Proceedings of the European Chapter of the Association for Computational Linguistics (EACL 2017)
Publisher	Association for Computational Linguistics
Pages	480-484
Number of pages	5
ISBN (Print)	978-1-945626-34-0
Publication status	Published - 7 Apr 2017
Event	The 15th Conference of the European Chapter of the Association for Computational Linguistics - Valencia, Spain Duration: 3 Apr 2017 → 7 Apr 2017

Conference

Conference	The 15th Conference of the European Chapter of the Association for Computational Linguistics
Abbreviated title	EACL 2017
Country/Territory	Spain
City	Valencia
Period	3/04/17 → 7/04/17

Access to Document

E17-2077Final published version, 691 KBLicence: Creative Commons: Attribution (CC-BY)

http://aclweb.org/anthology/E17-2077Licence: Creative Commons: Attribution (CC-BY)

STAC - strategic conversation
Lascarides, A. (Principal Investigator)
EU government bodies
1/06/11 → 31/05/17
Project: Research

Cite this

Keizer, S., Guhe, M., Cuayáhuitl, H., Efstathiou, I., Engelbrecht, K.-P., Dobre, M., Lascarides, A., & Lemon, O. (2017). Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents. In Proceedings of the European Chapter of the Association for Computational Linguistics (EACL 2017) (pp. 480-484). Association for Computational Linguistics. http://aclweb.org/anthology/E17-2077

@inproceedings{9f754c834f164f07b1ac05b3f81ddc07,

title = "Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents",

abstract = "In this paper we present a comparative evaluation of various negotiation strategies within an online version of the game “Settlers of Catan”. The comparison is based on human subjects playing games against artificial game-playing agents ({\textquoteleft}bots{\textquoteright}) which implement different negotiation dialogue strategies, using a chat dialogue interface to negotiate trades. Our results suggest that a negotiation strategy that uses persuasion, as well as a strategy that is trained from data using Deep Reinforcement Learning, both lead to an improved win rate against humans, compared to previous rule-based and supervised learning baseline dialogue negotiators.",

author = "Simon Keizer and Markus Guhe and Heriberto Cuay{\'a}huitl and Ioannis Efstathiou and Klaus-Peter Engelbrecht and Mihai Dobre and Alexandra Lascarides and Oliver Lemon",

year = "2017",

month = apr,

day = "7",

language = "English",

isbn = "978-1-945626-34-0",

pages = "480--484",

booktitle = "Proceedings of the European Chapter of the Association for Computational Linguistics (EACL 2017)",

publisher = "Association for Computational Linguistics",

note = "The 15th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2017 ; Conference date: 03-04-2017 Through 07-04-2017",

}

Keizer, S, Guhe, M, Cuayáhuitl, H, Efstathiou, I, Engelbrecht, K-P, Dobre, M, Lascarides, A & Lemon, O 2017, Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents. in Proceedings of the European Chapter of the Association for Computational Linguistics (EACL 2017). Association for Computational Linguistics, pp. 480-484, The 15th Conference of the European Chapter of the Association for Computational Linguistics, Valencia, Spain, 3/04/17. <http://aclweb.org/anthology/E17-2077>

Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents. / Keizer, Simon; Guhe, Markus; Cuayáhuitl, Heriberto et al.
Proceedings of the European Chapter of the Association for Computational Linguistics (EACL 2017). Association for Computational Linguistics, 2017. p. 480-484.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents

AU - Keizer, Simon

AU - Guhe, Markus

AU - Cuayáhuitl, Heriberto

AU - Efstathiou, Ioannis

AU - Engelbrecht, Klaus-Peter

AU - Dobre, Mihai

AU - Lascarides, Alexandra

AU - Lemon, Oliver

PY - 2017/4/7

Y1 - 2017/4/7

N2 - In this paper we present a comparative evaluation of various negotiation strategies within an online version of the game “Settlers of Catan”. The comparison is based on human subjects playing games against artificial game-playing agents (‘bots’) which implement different negotiation dialogue strategies, using a chat dialogue interface to negotiate trades. Our results suggest that a negotiation strategy that uses persuasion, as well as a strategy that is trained from data using Deep Reinforcement Learning, both lead to an improved win rate against humans, compared to previous rule-based and supervised learning baseline dialogue negotiators.

AB - In this paper we present a comparative evaluation of various negotiation strategies within an online version of the game “Settlers of Catan”. The comparison is based on human subjects playing games against artificial game-playing agents (‘bots’) which implement different negotiation dialogue strategies, using a chat dialogue interface to negotiate trades. Our results suggest that a negotiation strategy that uses persuasion, as well as a strategy that is trained from data using Deep Reinforcement Learning, both lead to an improved win rate against humans, compared to previous rule-based and supervised learning baseline dialogue negotiators.

M3 - Conference contribution

SN - 978-1-945626-34-0

SP - 480

EP - 484

BT - Proceedings of the European Chapter of the Association for Computational Linguistics (EACL 2017)

PB - Association for Computational Linguistics

T2 - The 15th Conference of the European Chapter of the Association for Computational Linguistics

Y2 - 3 April 2017 through 7 April 2017

ER -

Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents

Abstract

Conference

Access to Document

Fingerprint

Projects

STAC - strategic conversation

Cite this