Text Generation from Discourse Representation Structures

Jiangming Liu; Shay B. Cohen; Mirella Lapata

doi:10.18653/v1/2021.naacl-main.35

Text Generation from Discourse Representation Structures

Jiangming Liu, Shay B. Cohen, Mirella Lapata

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

Abstract

We propose neural models to generate text from formal meaning representations based on Discourse Representation Structures (DRSs). DRSs are document-level representations which encode rich semantic detail pertaining to rhetorical relations, presupposition, and co-reference within and across sentences. We formalize the task of neural DRS-to-text generation and provide modeling solutions for the problems of condition ordering and variable naming which render generation from DRSs non-trivial. Our generator relies on a novel sibling treeLSTM model which is able to accurately represent DRS structures and is more generally suited to trees with wide branches. We achieve competitive performance (59.48 BLEU) on the GMB benchmark against several strong baselines.

Original language	English
Title of host publication	Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Place of Publication	Online
Publisher	Association for Computational Linguistics
Pages	397-415
Number of pages	19
ISBN (Print)	978-1-954085-46-6
DOIs	https://doi.org/10.18653/v1/2021.naacl-main.35
Publication status	Published - 6 Jun 2021
Event	2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics - Duration: 6 Jun 2021 → 11 Jun 2021 https://2021.naacl.org/

Conference

Conference	2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics
Abbreviated title	NAACL 2021
Period	6/06/21 → 11/06/21
Internet address	https://2021.naacl.org/

Access to Document

10.18653/v1/2021.naacl-main.35Licence: Creative Commons: Attribution (CC-BY)

Text Generation_LIU_DOA06062021_VOR_CC-BYFinal published version, 521 KBLicence: Creative Commons: Attribution (CC-BY)

https://www.aclweb.org/anthology/2021.naacl-main.35Licence: Creative Commons: Attribution (CC-BY)

TransModal: Translating from Multiple Modalities into Text
Lapata, M.
EU government bodies
1/09/16 → 31/08/22
Project: Research
SUMMA - Scalable Understanding of Mulitingual Media
Renals, S., Birch-Mayne, A. & Cohen, S.
EU government bodies
1/02/16 → 31/01/19
Project: Research

Cite this

@inproceedings{135652fe1b20442da62c64693be6d773,

title = "Text Generation from Discourse Representation Structures",

abstract = "We propose neural models to generate text from formal meaning representations based on Discourse Representation Structures (DRSs). DRSs are document-level representations which encode rich semantic detail pertaining to rhetorical relations, presupposition, and co-reference within and across sentences. We formalize the task of neural DRS-to-text generation and provide modeling solutions for the problems of condition ordering and variable naming which render generation from DRSs non-trivial. Our generator relies on a novel sibling treeLSTM model which is able to accurately represent DRS structures and is more generally suited to trees with wide branches. We achieve competitive performance (59.48 BLEU) on the GMB benchmark against several strong baselines.",

author = "Jiangming Liu and Cohen, {Shay B.} and Mirella Lapata",

year = "2021",

month = jun,

day = "6",

doi = "10.18653/v1/2021.naacl-main.35",

language = "English",

isbn = "978-1-954085-46-6",

pages = "397--415",

booktitle = "Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies",

publisher = "Association for Computational Linguistics",

note = "2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics, NAACL 2021 ; Conference date: 06-06-2021 Through 11-06-2021",

url = "https://2021.naacl.org/",

}

Liu, J, Cohen, SB & Lapata, M 2021, Text Generation from Discourse Representation Structures. in Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Association for Computational Linguistics, Online, pp. 397-415, 2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics, 6/06/21. https://doi.org/10.18653/v1/2021.naacl-main.35

Text Generation from Discourse Representation Structures. / Liu, Jiangming; Cohen, Shay B.; Lapata, Mirella.
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Online: Association for Computational Linguistics, 2021. p. 397-415.

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

TY - GEN

T1 - Text Generation from Discourse Representation Structures

AU - Liu, Jiangming

AU - Cohen, Shay B.

AU - Lapata, Mirella

PY - 2021/6/6

Y1 - 2021/6/6

N2 - We propose neural models to generate text from formal meaning representations based on Discourse Representation Structures (DRSs). DRSs are document-level representations which encode rich semantic detail pertaining to rhetorical relations, presupposition, and co-reference within and across sentences. We formalize the task of neural DRS-to-text generation and provide modeling solutions for the problems of condition ordering and variable naming which render generation from DRSs non-trivial. Our generator relies on a novel sibling treeLSTM model which is able to accurately represent DRS structures and is more generally suited to trees with wide branches. We achieve competitive performance (59.48 BLEU) on the GMB benchmark against several strong baselines.

AB - We propose neural models to generate text from formal meaning representations based on Discourse Representation Structures (DRSs). DRSs are document-level representations which encode rich semantic detail pertaining to rhetorical relations, presupposition, and co-reference within and across sentences. We formalize the task of neural DRS-to-text generation and provide modeling solutions for the problems of condition ordering and variable naming which render generation from DRSs non-trivial. Our generator relies on a novel sibling treeLSTM model which is able to accurately represent DRS structures and is more generally suited to trees with wide branches. We achieve competitive performance (59.48 BLEU) on the GMB benchmark against several strong baselines.

U2 - 10.18653/v1/2021.naacl-main.35

DO - 10.18653/v1/2021.naacl-main.35

M3 - Conference contribution

SN - 978-1-954085-46-6

SP - 397

EP - 415

BT - Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

PB - Association for Computational Linguistics

CY - Online

T2 - 2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics

Y2 - 6 June 2021 through 11 June 2021

ER -

Text Generation from Discourse Representation Structures

Abstract

Conference

Access to Document

Fingerprint

Projects

TransModal: Translating from Multiple Modalities into Text

SUMMA - Scalable Understanding of Mulitingual Media

Cite this