Multitask parsing across semantic representations

Daniel Hershcovich; Omri Abend; Ari Rappoport

doi:10.18653/v1/p18-1035

Multitask parsing across semantic representations

Daniel Hershcovich, Omri Abend, Ari Rappoport

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

51 Scopus citations

Abstract

The ability to consolidate information of different types is at the core of intelligence, and has tremendous practical value in allowing learning for one task to benefit from generalizations learned for others. In this paper we tackle the challenging task of improving semantic parsing performance, taking UCCA parsing as a test case, and AMR, SDP and Universal Dependencies (UD) parsing as auxiliary tasks. We experiment on three languages, using a uniform transition-based system and learning architecture for all parsing tasks. Despite notable conceptual, formal and domain differences, we show that multitask learning significantly improves UCCA parsing in both in-domain and out-of-domain settings. Our code is publicly available.

Original language	American English
Title of host publication	ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)
Publisher	Association for Computational Linguistics (ACL)
Pages	373-385
Number of pages	13
ISBN (Electronic)	9781948087322
DOIs	https://doi.org/10.18653/v1/p18-1035
State	Published - 2018
Event	56th Annual Meeting of the Association for Computational Linguistics, ACL 2018 - Melbourne, Australia Duration: 15 Jul 2018 → 20 Jul 2018

Publication series

Name	ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)
Volume	1

Conference

Conference	56th Annual Meeting of the Association for Computational Linguistics, ACL 2018
Country/Territory	Australia
City	Melbourne
Period	15/07/18 → 20/07/18

Bibliographical note

Funding Information:
This work was supported by the Israel Science Foundation (grant no. 929/17), by the HUJI Cy-ber Security Research Center in conjunction with the Israel National Cyber Bureau in the Prime Minister’s Office, and by the Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI). The first author was supported by a fellowship from the Edmond and Lily Safra Center for Brain Sciences. We thank Roi Reichart, Rotem Dror and the anonymous reviewers for their helpful comments.

Publisher Copyright:
© 2018 Association for Computational Linguistics

Access to Document

10.18653/v1/p18-1035

Cite this

Hershcovich, D., Abend, O., & Rappoport, A. (2018). Multitask parsing across semantic representations. In ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers) (pp. 373-385). (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers); Vol. 1). Association for Computational Linguistics (ACL). https://doi.org/10.18653/v1/p18-1035

Hershcovich, Daniel ; Abend, Omri ; Rappoport, Ari. / Multitask parsing across semantic representations. ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). Association for Computational Linguistics (ACL), 2018. pp. 373-385 (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)).

@inproceedings{781ff04b02e747ef92b704e5d1e43467,

title = "Multitask parsing across semantic representations",

abstract = "The ability to consolidate information of different types is at the core of intelligence, and has tremendous practical value in allowing learning for one task to benefit from generalizations learned for others. In this paper we tackle the challenging task of improving semantic parsing performance, taking UCCA parsing as a test case, and AMR, SDP and Universal Dependencies (UD) parsing as auxiliary tasks. We experiment on three languages, using a uniform transition-based system and learning architecture for all parsing tasks. Despite notable conceptual, formal and domain differences, we show that multitask learning significantly improves UCCA parsing in both in-domain and out-of-domain settings. Our code is publicly available.",

author = "Daniel Hershcovich and Omri Abend and Ari Rappoport",

note = "Funding Information: This work was supported by the Israel Science Foundation (grant no. 929/17), by the HUJI Cy-ber Security Research Center in conjunction with the Israel National Cyber Bureau in the Prime Minister{\textquoteright}s Office, and by the Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI). The first author was supported by a fellowship from the Edmond and Lily Safra Center for Brain Sciences. We thank Roi Reichart, Rotem Dror and the anonymous reviewers for their helpful comments. Publisher Copyright: {\textcopyright} 2018 Association for Computational Linguistics; 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018 ; Conference date: 15-07-2018 Through 20-07-2018",

year = "2018",

doi = "10.18653/v1/p18-1035",

language = "American English",

series = "ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)",

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

pages = "373--385",

booktitle = "ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)",

address = "United States",

}

Hershcovich, D, Abend, O & Rappoport, A 2018, Multitask parsing across semantic representations. in ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers), vol. 1, Association for Computational Linguistics (ACL), pp. 373-385, 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018, Melbourne, Australia, 15/07/18. https://doi.org/10.18653/v1/p18-1035

Multitask parsing across semantic representations. / Hershcovich, Daniel; Abend, Omri ; Rappoport, Ari.
ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). Association for Computational Linguistics (ACL), 2018. p. 373-385 (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers); Vol. 1).

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

TY - GEN

T1 - Multitask parsing across semantic representations

AU - Hershcovich, Daniel

AU - Abend, Omri

AU - Rappoport, Ari

N1 - Funding Information: This work was supported by the Israel Science Foundation (grant no. 929/17), by the HUJI Cy-ber Security Research Center in conjunction with the Israel National Cyber Bureau in the Prime Minister’s Office, and by the Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI). The first author was supported by a fellowship from the Edmond and Lily Safra Center for Brain Sciences. We thank Roi Reichart, Rotem Dror and the anonymous reviewers for their helpful comments. Publisher Copyright: © 2018 Association for Computational Linguistics

PY - 2018

Y1 - 2018

N2 - The ability to consolidate information of different types is at the core of intelligence, and has tremendous practical value in allowing learning for one task to benefit from generalizations learned for others. In this paper we tackle the challenging task of improving semantic parsing performance, taking UCCA parsing as a test case, and AMR, SDP and Universal Dependencies (UD) parsing as auxiliary tasks. We experiment on three languages, using a uniform transition-based system and learning architecture for all parsing tasks. Despite notable conceptual, formal and domain differences, we show that multitask learning significantly improves UCCA parsing in both in-domain and out-of-domain settings. Our code is publicly available.

AB - The ability to consolidate information of different types is at the core of intelligence, and has tremendous practical value in allowing learning for one task to benefit from generalizations learned for others. In this paper we tackle the challenging task of improving semantic parsing performance, taking UCCA parsing as a test case, and AMR, SDP and Universal Dependencies (UD) parsing as auxiliary tasks. We experiment on three languages, using a uniform transition-based system and learning architecture for all parsing tasks. Despite notable conceptual, formal and domain differences, we show that multitask learning significantly improves UCCA parsing in both in-domain and out-of-domain settings. Our code is publicly available.

UR - http://www.scopus.com/inward/record.url?scp=85059485613&partnerID=8YFLogxK

U2 - 10.18653/v1/p18-1035

DO - 10.18653/v1/p18-1035

M3 - Conference contribution

AN - SCOPUS:85059485613

T3 - ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)

SP - 373

EP - 385

BT - ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)

PB - Association for Computational Linguistics (ACL)

T2 - 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018

Y2 - 15 July 2018 through 20 July 2018

ER -

Hershcovich D, Abend O , Rappoport A. Multitask parsing across semantic representations. In ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers). Association for Computational Linguistics (ACL). 2018. p. 373-385. (ACL 2018 - 56th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers)). doi: 10.18653/v1/p18-1035

Multitask parsing across semantic representations

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this