Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling

Itay Harel; Hagai Taitelbaum; Idan Szpektor; Oren Kurland

doi:10.1007/978-3-031-28244-7_27

Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling

Itay Harel, Hagai Taitelbaum, Idan Szpektor, Oren Kurland

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

Abstract

We address the task of retrieving sentences for an open domain dialogue that contain information useful for generating the next turn. We propose several novel neural retrieval architectures based on dual contextual modeling: the dialogue context and the context of the sentence in its ambient document. The architectures utilize contextualized language models (BERT), fine-tuned on a large-scale dataset constructed from Reddit. We evaluate the models using a recently published dataset. The performance of our most effective model is substantially superior to that of strong baselines.

Original language	English
Title of host publication	Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings
Editors	Jaap Kamps, Lorraine Goeuriot, Fabio Crestani, Maria Maistro, Hideo Joho, Brian Davis, Cathal Gurrin, Annalina Caputo, Udo Kruschwitz
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	426-440
Number of pages	15
ISBN (Print)	9783031282430
DOIs	https://doi.org/10.1007/978-3-031-28244-7_27
State	Published - 2023
Externally published	Yes
Event	45th European Conference on Information Retrieval, ECIR 2023 - Dublin, Ireland Duration: 2 Apr 2023 → 6 Apr 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13980 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	45th European Conference on Information Retrieval, ECIR 2023
Country/Territory	Ireland
City	Dublin
Period	2/04/23 → 6/04/23

Bibliographical note

Funding Information:
Acknowledgements. We thank the reviewers for their comments. This work was supported in part by a grant from Google.

Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Keywords

Dialogue retrieval
Open domain dialogue
Sentence retrieval

Access to Document

10.1007/978-3-031-28244-7_27

Cite this

Harel, I., Taitelbaum, H., Szpektor, I., & Kurland, O. (2023). Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling. In J. Kamps, L. Goeuriot, F. Crestani, M. Maistro, H. Joho, B. Davis, C. Gurrin, A. Caputo, & U. Kruschwitz (Eds.), Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings (pp. 426-440). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13980 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-28244-7_27

Harel, Itay ; Taitelbaum, Hagai ; Szpektor, Idan et al. / Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling. Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings. editor / Jaap Kamps ; Lorraine Goeuriot ; Fabio Crestani ; Maria Maistro ; Hideo Joho ; Brian Davis ; Cathal Gurrin ; Annalina Caputo ; Udo Kruschwitz. Springer Science and Business Media Deutschland GmbH, 2023. pp. 426-440 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{d475b6a6ee7d4021a39e2fd954f75a65,

title = "Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling",

abstract = "We address the task of retrieving sentences for an open domain dialogue that contain information useful for generating the next turn. We propose several novel neural retrieval architectures based on dual contextual modeling: the dialogue context and the context of the sentence in its ambient document. The architectures utilize contextualized language models (BERT), fine-tuned on a large-scale dataset constructed from Reddit. We evaluate the models using a recently published dataset. The performance of our most effective model is substantially superior to that of strong baselines.",

keywords = "Dialogue retrieval, Open domain dialogue, Sentence retrieval",

author = "Itay Harel and Hagai Taitelbaum and Idan Szpektor and Oren Kurland",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 45th European Conference on Information Retrieval, ECIR 2023 ; Conference date: 02-04-2023 Through 06-04-2023",

year = "2023",

doi = "10.1007/978-3-031-28244-7_27",

language = "אנגלית",

isbn = "9783031282430",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "426--440",

editor = "Jaap Kamps and Lorraine Goeuriot and Fabio Crestani and Maria Maistro and Hideo Joho and Brian Davis and Cathal Gurrin and Annalina Caputo and Udo Kruschwitz",

booktitle = "Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings",

address = "גרמניה",

}

Harel, I, Taitelbaum, H, Szpektor, I & Kurland, O 2023, Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling. in J Kamps, L Goeuriot, F Crestani, M Maistro, H Joho, B Davis, C Gurrin, A Caputo & U Kruschwitz (eds), Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13980 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 426-440, 45th European Conference on Information Retrieval, ECIR 2023, Dublin, Ireland, 2/04/23. https://doi.org/10.1007/978-3-031-28244-7_27

Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling. / Harel, Itay; Taitelbaum, Hagai; Szpektor, Idan et al.
Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings. ed. / Jaap Kamps; Lorraine Goeuriot; Fabio Crestani; Maria Maistro; Hideo Joho; Brian Davis; Cathal Gurrin; Annalina Caputo; Udo Kruschwitz. Springer Science and Business Media Deutschland GmbH, 2023. p. 426-440 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13980 LNCS).

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

TY - GEN

T1 - Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling

AU - Harel, Itay

AU - Taitelbaum, Hagai

AU - Szpektor, Idan

AU - Kurland, Oren

PY - 2023

Y1 - 2023

N2 - We address the task of retrieving sentences for an open domain dialogue that contain information useful for generating the next turn. We propose several novel neural retrieval architectures based on dual contextual modeling: the dialogue context and the context of the sentence in its ambient document. The architectures utilize contextualized language models (BERT), fine-tuned on a large-scale dataset constructed from Reddit. We evaluate the models using a recently published dataset. The performance of our most effective model is substantially superior to that of strong baselines.

AB - We address the task of retrieving sentences for an open domain dialogue that contain information useful for generating the next turn. We propose several novel neural retrieval architectures based on dual contextual modeling: the dialogue context and the context of the sentence in its ambient document. The architectures utilize contextualized language models (BERT), fine-tuned on a large-scale dataset constructed from Reddit. We evaluate the models using a recently published dataset. The performance of our most effective model is substantially superior to that of strong baselines.

KW - Dialogue retrieval

KW - Open domain dialogue

KW - Sentence retrieval

UR - http://www.scopus.com/inward/record.url?scp=85151150544&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-28244-7_27

DO - 10.1007/978-3-031-28244-7_27

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:85151150544

SN - 9783031282430

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 426

EP - 440

BT - Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings

A2 - Kamps, Jaap

A2 - Goeuriot, Lorraine

A2 - Crestani, Fabio

A2 - Maistro, Maria

A2 - Joho, Hideo

A2 - Davis, Brian

A2 - Gurrin, Cathal

A2 - Caputo, Annalina

A2 - Kruschwitz, Udo

PB - Springer Science and Business Media Deutschland GmbH

T2 - 45th European Conference on Information Retrieval, ECIR 2023

Y2 - 2 April 2023 through 6 April 2023

ER -

Harel I, Taitelbaum H, Szpektor I, Kurland O. Sentence Retrieval for Open-Ended Dialogue Using Dual Contextual Modeling. In Kamps J, Goeuriot L, Crestani F, Maistro M, Joho H, Davis B, Gurrin C, Caputo A, Kruschwitz U, editors, Advances in Information Retrieval - 45th European Conference on Information Retrieval, ECIR 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 426-440. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-28244-7_27