Ellipsis Resolution as Question Answering: An Evaluation

Rahul Aralikatte; Matthew Lamm; Daniel Hardt; Anders Søgaard

doi:10.18653/v1/2021.eacl-main.68

Ellipsis Resolution as Question Answering: An Evaluation

Rahul Aralikatte
, Matthew Lamm
, Daniel Hardt
, Anders Søgaard

Department of Management, Society and Communication

University of Copenhagen
Stanford University

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

169 Downloads (Pure)

Abstract

Most, if not all forms of ellipsis (e.g., ‘so does Mary’) are similar to reading comprehension questions (‘what does Mary do’), in that in order to resolve them, we need to identify an appropriate text span in the preceding discourse. Following this observation, we present an alternative approach for English ellipsis resolution relying on architectures developed for question answering (QA). We present both single-task models, and joint models trained on auxiliary QA and coreference resolution datasets, clearly outperforming the current state of the art for Sluice Ellipsis (from 70.00 to 86.01 F1) and Verb Phrase Ellipsis (from 72.89 to 78.66 F1).

Original language	English
Title of host publication	Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics
Editors	Paola Merlo, Jorg Tiedemann, Reut Tsarfaty
Number of pages	8
Place of Publication	Stroudsburg, PA
Publisher	Association for Computational Linguistics
Publication date	2021
Pages	810-817
ISBN (Electronic)	9781954085022
DOIs	https://doi.org/10.18653/v1/2021.eacl-main.68
Publication status	Published - 2021
Event	16th Conference of the European Chapter of the Association for Computational Linguistics (EACL) - Online, WWW Duration: 19 Apr 2021 → 23 Apr 2021 Conference number: 16

Conference

Conference	16th Conference of the European Chapter of the Association for Computational Linguistics (EACL)
Number	16
Location	Online
Country/Territory	WWW
Period	19/04/2021 → 23/04/2021

Access to Document

10.18653/v1/2021.eacl-main.68Licence: CC BY

rahul_aralikatte_et_al_ellipsis_resolution_as_question_answering_an_evaluation_publishersversionFinal published version, 444 KBLicence: CC BY

Handle.net

Persistent link

Cite this

Aralikatte, R., Lamm, M., Hardt, D., & Søgaard, A. (2021). Ellipsis Resolution as Question Answering: An Evaluation. In P. Merlo, J. Tiedemann, & R. Tsarfaty (Eds.), Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics (pp. 810-817). Association for Computational Linguistics. https://doi.org/10.18653/v1/2021.eacl-main.68

@inproceedings{6cd9bb1f03d34ea6b97afa66bd9068fd,

title = "Ellipsis Resolution as Question Answering: An Evaluation",

abstract = "Most, if not all forms of ellipsis (e.g., {\textquoteleft}so does Mary{\textquoteright}) are similar to reading comprehension questions ({\textquoteleft}what does Mary do{\textquoteright}), in that in order to resolve them, we need to identify an appropriate text span in the preceding discourse. Following this observation, we present an alternative approach for English ellipsis resolution relying on architectures developed for question answering (QA). We present both single-task models, and joint models trained on auxiliary QA and coreference resolution datasets, clearly outperforming the current state of the art for Sluice Ellipsis (from 70.00 to 86.01 F1) and Verb Phrase Ellipsis (from 72.89 to 78.66 F1).",

author = "Rahul Aralikatte and Matthew Lamm and Daniel Hardt and Anders S{\o}gaard",

year = "2021",

doi = "10.18653/v1/2021.eacl-main.68",

language = "English",

pages = "810--817",

editor = "Paola Merlo and Jorg Tiedemann and Tsarfaty, \{Reut \}",

booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics",

address = "United States",

note = "16th Conference of the European Chapter of the Association for Computational Linguistics (EACL), EACL ; Conference date: 19-04-2021 Through 23-04-2021",

}

Aralikatte, R, Lamm, M, Hardt, D & Søgaard, A 2021, Ellipsis Resolution as Question Answering: An Evaluation. in P Merlo, J Tiedemann & R Tsarfaty (eds), Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics. Association for Computational Linguistics, Stroudsburg, PA, pp. 810-817, 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL), WWW, 19/04/2021. https://doi.org/10.18653/v1/2021.eacl-main.68

Ellipsis Resolution as Question Answering: An Evaluation. / Aralikatte, Rahul; Lamm, Matthew; Hardt, Daniel et al.
Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics. ed. / Paola Merlo; Jorg Tiedemann; Reut Tsarfaty. Stroudsburg, PA: Association for Computational Linguistics, 2021. p. 810-817.

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

TY - GEN

T1 - Ellipsis Resolution as Question Answering

T2 - 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL)

AU - Aralikatte, Rahul

AU - Lamm, Matthew

AU - Hardt, Daniel

AU - Søgaard, Anders

N1 - Conference code: 16

PY - 2021

Y1 - 2021

N2 - Most, if not all forms of ellipsis (e.g., ‘so does Mary’) are similar to reading comprehension questions (‘what does Mary do’), in that in order to resolve them, we need to identify an appropriate text span in the preceding discourse. Following this observation, we present an alternative approach for English ellipsis resolution relying on architectures developed for question answering (QA). We present both single-task models, and joint models trained on auxiliary QA and coreference resolution datasets, clearly outperforming the current state of the art for Sluice Ellipsis (from 70.00 to 86.01 F1) and Verb Phrase Ellipsis (from 72.89 to 78.66 F1).

AB - Most, if not all forms of ellipsis (e.g., ‘so does Mary’) are similar to reading comprehension questions (‘what does Mary do’), in that in order to resolve them, we need to identify an appropriate text span in the preceding discourse. Following this observation, we present an alternative approach for English ellipsis resolution relying on architectures developed for question answering (QA). We present both single-task models, and joint models trained on auxiliary QA and coreference resolution datasets, clearly outperforming the current state of the art for Sluice Ellipsis (from 70.00 to 86.01 F1) and Verb Phrase Ellipsis (from 72.89 to 78.66 F1).

U2 - 10.18653/v1/2021.eacl-main.68

DO - 10.18653/v1/2021.eacl-main.68

M3 - Article in proceedings

SP - 810

EP - 817

BT - Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics

A2 - Merlo, Paola

A2 - Tiedemann, Jorg

A2 - Tsarfaty, Reut

PB - Association for Computational Linguistics

CY - Stroudsburg, PA

Y2 - 19 April 2021 through 23 April 2021

ER -