A mention-based system for revision requirements detection

Ahmed Ruby; Christian Hardmeier; Sara Stymne

A mention-based system for revision requirements detection

Ahmed Ruby
, Christian Hardmeier
, Sara Stymne

NLPnorth

Uppsala University

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

Abstract

Exploring aspects of sentential meaning that are implicit or underspecified in context is important for sentence understanding. In this paper, we propose a novel architecture based on mentions for revision requirements detection. The goal is to improve understandability, addressing some types of revisions, especially for the Replaced Pronoun type. We show that our mention-based system can predict replaced pronouns well on the mentionlevel. However, our combined sentence-level system does not improve on the sentence-level BERT baseline. We also present additional contrastive systems, and show results for each type of edit.

Originalsprog	Engelsk
Titel	Proceedings of the First Workshop on Understanding Implicit and Underspecified Language (UnImplicit)
Forlag	Association for Computational Linguistics
Publikationsdato	2021
Sider	58-63
Status	Udgivet - 2021
Begivenhed	Workshop on Understanding Implicit and Underspecified Language - VIRTUAL Varighed: 5 aug. 2021 → 6 aug. 2021 Konferencens nummer: 1

Workshop

Workshop	Workshop on Understanding Implicit and Underspecified Language
Nummer	1
By	VIRTUAL
Periode	05/08/2021 → 06/08/2021

Emneord

Sentential meaning
Contextual underspecification
Revision requirements detection
Mention-based architecture
Replaced Pronoun detection

Adgang til dokumentet

2021.unimplicit-1.7Forlagets udgivne version, 170 KBLicens: CC BY

https://aclanthology.org/2021.unimplicit-1.7.pdfLicens: CC BY

Citationsformater

@inproceedings{130119f2e6f54b18a96b6380f9261ac9,

title = "A mention-based system for revision requirements detection",

abstract = "Exploring aspects of sentential meaning that are implicit or underspecified in context is important for sentence understanding. In this paper, we propose a novel architecture based on mentions for revision requirements detection. The goal is to improve understandability, addressing some types of revisions, especially for the Replaced Pronoun type. We show that our mention-based system can predict replaced pronouns well on the mentionlevel. However, our combined sentence-level system does not improve on the sentence-level BERT baseline. We also present additional contrastive systems, and show results for each type of edit.",

keywords = "Sentential meaning, Contextual underspecification, Revision requirements detection, Mention-based architecture, Replaced Pronoun detection, Sentential meaning, Contextual underspecification, Revision requirements detection, Mention-based architecture, Replaced Pronoun detection",

author = "Ahmed Ruby and Christian Hardmeier and Sara Stymne",

year = "2021",

language = "English",

pages = "58--63",

booktitle = "Proceedings of the First Workshop on Understanding Implicit and Underspecified Language (UnImplicit)",

publisher = "Association for Computational Linguistics",

address = "United States",

note = "Workshop on Understanding Implicit and Underspecified Language, UnImplicit ; Conference date: 05-08-2021 Through 06-08-2021",

}

Ruby, A, Hardmeier, C & Stymne, S 2021, A mention-based system for revision requirements detection. i Proceedings of the First Workshop on Understanding Implicit and Underspecified Language (UnImplicit). Association for Computational Linguistics, s. 58-63, Workshop on Understanding Implicit and Underspecified Language, VIRTUAL, 05/08/2021. <https://aclanthology.org/2021.unimplicit-1.7.pdf>

A mention-based system for revision requirements detection. / Ruby, Ahmed; Hardmeier, Christian; Stymne, Sara.
Proceedings of the First Workshop on Understanding Implicit and Underspecified Language (UnImplicit). Association for Computational Linguistics, 2021. s. 58-63.

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

TY - GEN

T1 - A mention-based system for revision requirements detection

AU - Ruby, Ahmed

AU - Hardmeier, Christian

AU - Stymne, Sara

N1 - Conference code: 1

PY - 2021

Y1 - 2021

N2 - Exploring aspects of sentential meaning that are implicit or underspecified in context is important for sentence understanding. In this paper, we propose a novel architecture based on mentions for revision requirements detection. The goal is to improve understandability, addressing some types of revisions, especially for the Replaced Pronoun type. We show that our mention-based system can predict replaced pronouns well on the mentionlevel. However, our combined sentence-level system does not improve on the sentence-level BERT baseline. We also present additional contrastive systems, and show results for each type of edit.

AB - Exploring aspects of sentential meaning that are implicit or underspecified in context is important for sentence understanding. In this paper, we propose a novel architecture based on mentions for revision requirements detection. The goal is to improve understandability, addressing some types of revisions, especially for the Replaced Pronoun type. We show that our mention-based system can predict replaced pronouns well on the mentionlevel. However, our combined sentence-level system does not improve on the sentence-level BERT baseline. We also present additional contrastive systems, and show results for each type of edit.

KW - Sentential meaning

KW - Contextual underspecification

KW - Revision requirements detection

KW - Mention-based architecture

KW - Replaced Pronoun detection

KW - Sentential meaning

KW - Contextual underspecification

KW - Revision requirements detection

KW - Mention-based architecture

KW - Replaced Pronoun detection

M3 - Article in proceedings

SP - 58

EP - 63

BT - Proceedings of the First Workshop on Understanding Implicit and Underspecified Language (UnImplicit)

PB - Association for Computational Linguistics

T2 - Workshop on Understanding Implicit and Underspecified Language

Y2 - 5 August 2021 through 6 August 2021

ER -

A mention-based system for revision requirements detection

Abstract

Workshop

Emneord

Adgang til dokumentet

Fingeraftryk

Citationsformater