Optimising natural language generation decision making for situated dialogue

Nina Dethlefs; Heriberto Cuayáhuitl; Jette Viethen

Optimising natural language generation decision making for situated dialogue

Nina Dethlefs^*, Heriberto Cuayáhuitl, Jette Viethen

^*Corresponding author for this work

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

14 Citations (Scopus)

Abstract

Natural language generators are faced with a multitude of different decisions during their generation process. We address the joint optimisation of navigation strategies and referring expressions in a situated setting with respect to task success and human-likeness. To this end, we present a novel, comprehensive framework that combines supervised learning, Hierarchical Reinforcement Learning and a hierarchical Information State. A human evaluation shows that our learnt instructions are rated similar to human instructions, and significantly better than the supervised learning baseline.

Original language	English
Title of host publication	Proceedings of the SIGDIAL 2011 Conference
Subtitle of host publication	The 12th Annual Meeting of the Special Interest Group on Discourse and Dialogue
Place of Publication	Stroudsburg, PA
Publisher	Association for Computational Linguistics (ACL)
Pages	78-87
Number of pages	10
ISBN (Print)	9781937284107
Publication status	Published - 2011
Event	12th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2011 - Portland, OR, United States Duration: 17 Jun 2011 → 18 Jun 2011

Other

Other	12th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2011
Country/Territory	United States
City	Portland, OR
Period	17/06/11 → 18/06/11

Cite this

@inproceedings{1862826179df45d5a534cda9eb7ab9d2,

title = "Optimising natural language generation decision making for situated dialogue",

abstract = "Natural language generators are faced with a multitude of different decisions during their generation process. We address the joint optimisation of navigation strategies and referring expressions in a situated setting with respect to task success and human-likeness. To this end, we present a novel, comprehensive framework that combines supervised learning, Hierarchical Reinforcement Learning and a hierarchical Information State. A human evaluation shows that our learnt instructions are rated similar to human instructions, and significantly better than the supervised learning baseline.",

author = "Nina Dethlefs and Heriberto Cuay{\'a}huitl and Jette Viethen",

year = "2011",

language = "English",

isbn = "9781937284107",

pages = "78--87",

booktitle = "Proceedings of the SIGDIAL 2011 Conference",

publisher = "Association for Computational Linguistics (ACL)",

note = "12th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2011 ; Conference date: 17-06-2011 Through 18-06-2011",

}

Dethlefs, N, Cuayáhuitl, H & Viethen, J 2011, Optimising natural language generation decision making for situated dialogue. in Proceedings of the SIGDIAL 2011 Conference: The 12th Annual Meeting of the Special Interest Group on Discourse and Dialogue. Association for Computational Linguistics (ACL), Stroudsburg, PA, pp. 78-87, 12th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2011, Portland, OR, United States, 17/06/11.

Optimising natural language generation decision making for situated dialogue. / Dethlefs, Nina; Cuayáhuitl, Heriberto; Viethen, Jette.
Proceedings of the SIGDIAL 2011 Conference: The 12th Annual Meeting of the Special Interest Group on Discourse and Dialogue. Stroudsburg, PA: Association for Computational Linguistics (ACL), 2011. p. 78-87.

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

TY - GEN

T1 - Optimising natural language generation decision making for situated dialogue

AU - Dethlefs, Nina

AU - Cuayáhuitl, Heriberto

AU - Viethen, Jette

PY - 2011

Y1 - 2011

N2 - Natural language generators are faced with a multitude of different decisions during their generation process. We address the joint optimisation of navigation strategies and referring expressions in a situated setting with respect to task success and human-likeness. To this end, we present a novel, comprehensive framework that combines supervised learning, Hierarchical Reinforcement Learning and a hierarchical Information State. A human evaluation shows that our learnt instructions are rated similar to human instructions, and significantly better than the supervised learning baseline.

AB - Natural language generators are faced with a multitude of different decisions during their generation process. We address the joint optimisation of navigation strategies and referring expressions in a situated setting with respect to task success and human-likeness. To this end, we present a novel, comprehensive framework that combines supervised learning, Hierarchical Reinforcement Learning and a hierarchical Information State. A human evaluation shows that our learnt instructions are rated similar to human instructions, and significantly better than the supervised learning baseline.

UR - http://www.scopus.com/inward/record.url?scp=84857730851&partnerID=8YFLogxK

M3 - Conference proceeding contribution

AN - SCOPUS:84857730851

SN - 9781937284107

SP - 78

EP - 87

BT - Proceedings of the SIGDIAL 2011 Conference

PB - Association for Computational Linguistics (ACL)

CY - Stroudsburg, PA

T2 - 12th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2011

Y2 - 17 June 2011 through 18 June 2011

ER -

Optimising natural language generation decision making for situated dialogue

Abstract

Other

Other files and links

Fingerprint

Cite this