Active learning for coreference resolution using discrete annotation

Belinda Z. Li; Gabriel Stanovsky; Luke Zettlemoyer

Active learning for coreference resolution using discrete annotation

Belinda Z. Li^*, Gabriel Stanovsky, Luke Zettlemoyer

^*Corresponding author for this work

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

19 Scopus citations

Abstract

We improve upon pairwise annotation for active learning in coreference resolution, by asking annotators to identify mention antecedents if a presented mention pair is deemed not coreferent. This simple modification, when combined with a novel mention clustering algorithm for selecting which examples to label, is much more efficient in terms of the performance obtained per annotation budget. In experiments with existing benchmark coreference datasets, we show that the signal from this additional question leads to significant performance gains per human-annotation hour. Future work can use our annotation protocol to effectively develop coreference models for new domains. Our code is publicly available.

Original language	English
Title of host publication	ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference
Publisher	Association for Computational Linguistics (ACL)
Pages	8320-8331
Number of pages	12
ISBN (Electronic)	9781952148255
State	Published - 2020
Externally published	Yes
Event	58th Annual Meeting of the Association for Computational Linguistics, ACL 2020 - Virtual, Online, United States Duration: 5 Jul 2020 → 10 Jul 2020

Publication series

Name	Proceedings of the Annual Meeting of the Association for Computational Linguistics
ISSN (Print)	0736-587X

Conference

Conference	58th Annual Meeting of the Association for Computational Linguistics, ACL 2020
Country/Territory	United States
City	Virtual, Online
Period	5/07/20 → 10/07/20

Bibliographical note

Publisher Copyright:
© 2020 Association for Computational Linguistics

Cite this

Li, B. Z., Stanovsky, G., & Zettlemoyer, L. (2020). Active learning for coreference resolution using discrete annotation. In ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (pp. 8320-8331). (Proceedings of the Annual Meeting of the Association for Computational Linguistics). Association for Computational Linguistics (ACL).

Li, Belinda Z. ; Stanovsky, Gabriel ; Zettlemoyer, Luke. / Active learning for coreference resolution using discrete annotation. ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference. Association for Computational Linguistics (ACL), 2020. pp. 8320-8331 (Proceedings of the Annual Meeting of the Association for Computational Linguistics).

@inproceedings{7f3e1b7876a14e9fb4fa4f778593d58c,

title = "Active learning for coreference resolution using discrete annotation",

abstract = "We improve upon pairwise annotation for active learning in coreference resolution, by asking annotators to identify mention antecedents if a presented mention pair is deemed not coreferent. This simple modification, when combined with a novel mention clustering algorithm for selecting which examples to label, is much more efficient in terms of the performance obtained per annotation budget. In experiments with existing benchmark coreference datasets, we show that the signal from this additional question leads to significant performance gains per human-annotation hour. Future work can use our annotation protocol to effectively develop coreference models for new domains. Our code is publicly available.",

author = "Li, {Belinda Z.} and Gabriel Stanovsky and Luke Zettlemoyer",

note = "Publisher Copyright: {\textcopyright} 2020 Association for Computational Linguistics; 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020 ; Conference date: 05-07-2020 Through 10-07-2020",

year = "2020",

language = "אנגלית",

series = "Proceedings of the Annual Meeting of the Association for Computational Linguistics",

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

pages = "8320--8331",

booktitle = "ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference",

address = "ארצות הברית",

}

Li, BZ, Stanovsky, G & Zettlemoyer, L 2020, Active learning for coreference resolution using discrete annotation. in ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference. Proceedings of the Annual Meeting of the Association for Computational Linguistics, Association for Computational Linguistics (ACL), pp. 8320-8331, 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020, Virtual, Online, United States, 5/07/20.

Active learning for coreference resolution using discrete annotation. / Li, Belinda Z.; Stanovsky, Gabriel; Zettlemoyer, Luke.
ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference. Association for Computational Linguistics (ACL), 2020. p. 8320-8331 (Proceedings of the Annual Meeting of the Association for Computational Linguistics).

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

TY - GEN

T1 - Active learning for coreference resolution using discrete annotation

AU - Li, Belinda Z.

AU - Stanovsky, Gabriel

AU - Zettlemoyer, Luke

PY - 2020

Y1 - 2020

N2 - We improve upon pairwise annotation for active learning in coreference resolution, by asking annotators to identify mention antecedents if a presented mention pair is deemed not coreferent. This simple modification, when combined with a novel mention clustering algorithm for selecting which examples to label, is much more efficient in terms of the performance obtained per annotation budget. In experiments with existing benchmark coreference datasets, we show that the signal from this additional question leads to significant performance gains per human-annotation hour. Future work can use our annotation protocol to effectively develop coreference models for new domains. Our code is publicly available.

AB - We improve upon pairwise annotation for active learning in coreference resolution, by asking annotators to identify mention antecedents if a presented mention pair is deemed not coreferent. This simple modification, when combined with a novel mention clustering algorithm for selecting which examples to label, is much more efficient in terms of the performance obtained per annotation budget. In experiments with existing benchmark coreference datasets, we show that the signal from this additional question leads to significant performance gains per human-annotation hour. Future work can use our annotation protocol to effectively develop coreference models for new domains. Our code is publicly available.

UR - http://www.scopus.com/inward/record.url?scp=85095538272&partnerID=8YFLogxK

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

AN - SCOPUS:85095538272

T3 - Proceedings of the Annual Meeting of the Association for Computational Linguistics

SP - 8320

EP - 8331

BT - ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference

PB - Association for Computational Linguistics (ACL)

T2 - 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020

Y2 - 5 July 2020 through 10 July 2020

ER -

Li BZ, Stanovsky G, Zettlemoyer L. Active learning for coreference resolution using discrete annotation. In ACL 2020 - 58th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference. Association for Computational Linguistics (ACL). 2020. p. 8320-8331. (Proceedings of the Annual Meeting of the Association for Computational Linguistics).

Active learning for coreference resolution using discrete annotation

Abstract

Publication series

Conference

Bibliographical note

Other files and links

Fingerprint

Cite this