Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11

Seokhwan Kim; Spandana Gella; Chao Zhao; Di Jin; Alexandros Papangelis; Behnam Hedayatnia; Yang Liu; Dilek Z Hakkani-Tur

Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11

Seokhwan Kim, Spandana Gella, Chao Zhao, Di Jin, Alexandros Papangelis, Behnam Hedayatnia, Yang Liu, Dilek Z Hakkani-Tur

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

Abstract

Conventional Task-oriented Dialogue (TOD) Systems rely on domain-specific APIs/DBs or external factual knowledge to create responses. In DSTC11 track 5, we aims to provide a new challenging task to accommodate subjective user requests (e.g.,''Is the WIFI reliable?'' or ``Does the restaurant have a good atmosphere?'' into TOD. We release a benchmark dataset, which contains subjective knowledge-seeking dialogue contexts and manually annotated responses that are grounded in subjective knowledge sources. The challenge track received a total of 48 entries from 14 participating teams.

Original language	English (US)
Title of host publication	Proceedings of The Eleventh Dialog System Technology Challenge
Editors	Yun-Nung Chen, Paul Crook, Michel Galley, Sarik Ghazarian, Chulaka Gunasekara, Raghav Gupta, Behnam Hedayatnia, Satwik Kottur, Seungwhan Moon, Chen Zhang
Place of Publication	Prague, Czech Republic
Publisher	Association for Computational Linguistics
Pages	274-281
Number of pages	8
State	Published - Sep 2023
Externally published	Yes

Online availability

https://aclanthology.org/2023.dstc-1.29

Library availability

Discover UIUC Full Text

Cite this

Kim, S., Gella, S., Zhao, C., Jin, D., Papangelis, A., Hedayatnia, B., Liu, Y., & Z Hakkani-Tur, D. (2023). Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11. In Y.-N. Chen, P. Crook, M. Galley, S. Ghazarian, C. Gunasekara, R. Gupta, B. Hedayatnia, S. Kottur, S. Moon, & C. Zhang (Eds.), Proceedings of The Eleventh Dialog System Technology Challenge (pp. 274-281). Association for Computational Linguistics. https://aclanthology.org/2023.dstc-1.29

Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11. / Kim, Seokhwan; Gella, Spandana; Zhao, Chao et al.
Proceedings of The Eleventh Dialog System Technology Challenge. ed. / Yun-Nung Chen; Paul Crook; Michel Galley; Sarik Ghazarian; Chulaka Gunasekara; Raghav Gupta; Behnam Hedayatnia; Satwik Kottur; Seungwhan Moon; Chen Zhang. Prague, Czech Republic: Association for Computational Linguistics, 2023. p. 274-281.

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

Kim, S, Gella, S, Zhao, C, Jin, D, Papangelis, A, Hedayatnia, B, Liu, Y & Z Hakkani-Tur, D 2023, Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11. in Y-N Chen, P Crook, M Galley, S Ghazarian, C Gunasekara, R Gupta, B Hedayatnia, S Kottur, S Moon & C Zhang (eds), Proceedings of The Eleventh Dialog System Technology Challenge. Association for Computational Linguistics, Prague, Czech Republic, pp. 274-281. <https://aclanthology.org/2023.dstc-1.29>

Kim S, Gella S, Zhao C, Jin D, Papangelis A, Hedayatnia B et al. Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11. In Chen YN, Crook P, Galley M, Ghazarian S, Gunasekara C, Gupta R, Hedayatnia B, Kottur S, Moon S, Zhang C, editors, Proceedings of The Eleventh Dialog System Technology Challenge. Prague, Czech Republic: Association for Computational Linguistics. 2023. p. 274-281

Kim, Seokhwan ; Gella, Spandana ; Zhao, Chao et al. / Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11. Proceedings of The Eleventh Dialog System Technology Challenge. editor / Yun-Nung Chen ; Paul Crook ; Michel Galley ; Sarik Ghazarian ; Chulaka Gunasekara ; Raghav Gupta ; Behnam Hedayatnia ; Satwik Kottur ; Seungwhan Moon ; Chen Zhang. Prague, Czech Republic : Association for Computational Linguistics, 2023. pp. 274-281

@inproceedings{7521ac1f1dba4b219f5e02e2122b5abf,

title = "Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11",

abstract = "Conventional Task-oriented Dialogue (TOD) Systems rely on domain-specific APIs/DBs or external factual knowledge to create responses. In DSTC11 track 5, we aims to provide a new challenging task to accommodate subjective user requests (e.g.,''Is the WIFI reliable?'' or ``Does the restaurant have a good atmosphere?'' into TOD. We release a benchmark dataset, which contains subjective knowledge-seeking dialogue contexts and manually annotated responses that are grounded in subjective knowledge sources. The challenge track received a total of 48 entries from 14 participating teams.",

author = "Seokhwan Kim and Spandana Gella and Chao Zhao and Di Jin and Alexandros Papangelis and Behnam Hedayatnia and Yang Liu and {Z Hakkani-Tur}, Dilek",

year = "2023",

month = sep,

language = "English (US)",

pages = "274--281",

editor = "Yun-Nung Chen and Paul Crook and Michel Galley and Sarik Ghazarian and Chulaka Gunasekara and Raghav Gupta and Behnam Hedayatnia and Satwik Kottur and Seungwhan Moon and Chen Zhang",

booktitle = "Proceedings of The Eleventh Dialog System Technology Challenge",

publisher = "Association for Computational Linguistics",

}

TY - GEN

T1 - Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11

AU - Kim, Seokhwan

AU - Gella, Spandana

AU - Zhao, Chao

AU - Jin, Di

AU - Papangelis, Alexandros

AU - Hedayatnia, Behnam

AU - Liu, Yang

AU - Z Hakkani-Tur, Dilek

PY - 2023/9

Y1 - 2023/9

N2 - Conventional Task-oriented Dialogue (TOD) Systems rely on domain-specific APIs/DBs or external factual knowledge to create responses. In DSTC11 track 5, we aims to provide a new challenging task to accommodate subjective user requests (e.g.,''Is the WIFI reliable?'' or ``Does the restaurant have a good atmosphere?'' into TOD. We release a benchmark dataset, which contains subjective knowledge-seeking dialogue contexts and manually annotated responses that are grounded in subjective knowledge sources. The challenge track received a total of 48 entries from 14 participating teams.

AB - Conventional Task-oriented Dialogue (TOD) Systems rely on domain-specific APIs/DBs or external factual knowledge to create responses. In DSTC11 track 5, we aims to provide a new challenging task to accommodate subjective user requests (e.g.,''Is the WIFI reliable?'' or ``Does the restaurant have a good atmosphere?'' into TOD. We release a benchmark dataset, which contains subjective knowledge-seeking dialogue contexts and manually annotated responses that are grounded in subjective knowledge sources. The challenge track received a total of 48 entries from 14 participating teams.

M3 - Conference contribution

SP - 274

EP - 281

BT - Proceedings of The Eleventh Dialog System Technology Challenge

A2 - Chen, Yun-Nung

A2 - Crook, Paul

A2 - Galley, Michel

A2 - Ghazarian, Sarik

A2 - Gunasekara, Chulaka

A2 - Gupta, Raghav

A2 - Hedayatnia, Behnam

A2 - Kottur, Satwik

A2 - Moon, Seungwhan

A2 - Zhang, Chen

PB - Association for Computational Linguistics

CY - Prague, Czech Republic

ER -

Task-Oriented Conversational Modeling with Subjective Knowledge Track in DSTC11

Abstract

Online availability

Library availability

Fingerprint

Cite this