Multi-Task Learning of Keyphrase Boundary Classification

Isabelle Augenstein; Anders Søgaard

doi:10.18653/v1/p17-2054

Multi-Task Learning of Keyphrase Boundary Classification

Isabelle Augenstein, Anders Søgaard

Datalogisk Institut

20 Citationer (Scopus)

Abstract

Keyphrase boundary classification (KBC) is the task of detecting keyphrases in scientific articles and labelling them with respect to predefined types. Although important in practice, this task is so far un-derexplored, partly due to the lack of labelled data. To overcome this, we explore several auxiliary tasks, including semantic super-sense tagging and identification of multi-word expressions, and cast the task as a multi-task learning problem with deep recurrent neural networks. Our multi-task models perform significantly better than previous state of the art approaches on two scientific KBC datasets, particularly for long keyphrases.

Originalsprog	Engelsk
Titel	Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics : (Short Papers)
Redaktører	Regina Barzilay, Min-Yen Kan
Antal sider	6
Vol/bind	2
Forlag	Association for Computational Linguistics
Publikationsdato	2017
Sider	341-346
ISBN (Trykt)	978-1-945626-76-0
DOI	https://doi.org/10.18653/v1/p17-2054
Status	Udgivet - 2017
Begivenhed	55th Annual Meeting of the Association for Computational Linguistics - Vancouver, Canada Varighed: 1 jul. 2017 → 1 jul. 2017

Konference

Konference	55th Annual Meeting of the Association for Computational Linguistics
Lokation	Vancouver, Canada
Periode	01/07/2017 → 01/07/2017

Adgang til dokumentet

10.18653/v1/p17-2054

P17-2054.pdfForlagets udgivne version, 439 KBLicens: CC BY

Citationsformater

Multi-Task Learning of Keyphrase Boundary Classification. / Augenstein, Isabelle ; Søgaard, Anders.
Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics : (Short Papers). red. / Regina Barzilay; Min-Yen Kan. Bind 2 Association for Computational Linguistics, 2017. s. 341-346.

Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › peer review

Augenstein, I & Søgaard, A 2017, Multi-Task Learning of Keyphrase Boundary Classification. i R Barzilay & M-Y Kan (red), Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics : (Short Papers). bind 2, Association for Computational Linguistics, s. 341-346, 55th Annual Meeting of the Association for Computational Linguistics , 01/07/2017. https://doi.org/10.18653/v1/p17-2054

@inproceedings{0fb1f2957ed742268aadd9e3594c4580,

title = "Multi-Task Learning of Keyphrase Boundary Classification",

abstract = "Keyphrase boundary classification (KBC) is the task of detecting keyphrases in scientific articles and labelling them with respect to predefined types. Although important in practice, this task is so far un-derexplored, partly due to the lack of labelled data. To overcome this, we explore several auxiliary tasks, including semantic super-sense tagging and identification of multi-word expressions, and cast the task as a multi-task learning problem with deep recurrent neural networks. Our multi-task models perform significantly better than previous state of the art approaches on two scientific KBC datasets, particularly for long keyphrases.",

author = "Isabelle Augenstein and Anders S{\o}gaard",

year = "2017",

doi = "10.18653/v1/p17-2054",

language = "English",

isbn = "978-1-945626-76-0",

volume = "2",

pages = "341--346",

editor = "Regina Barzilay and Min-Yen Kan",

booktitle = "Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics",

note = "Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) ; Conference date: 01-07-2017 Through 01-07-2017",

}

TY - GEN

T1 - Multi-Task Learning of Keyphrase Boundary Classification

AU - Augenstein, Isabelle

AU - Søgaard, Anders

PY - 2017

Y1 - 2017

N2 - Keyphrase boundary classification (KBC) is the task of detecting keyphrases in scientific articles and labelling them with respect to predefined types. Although important in practice, this task is so far un-derexplored, partly due to the lack of labelled data. To overcome this, we explore several auxiliary tasks, including semantic super-sense tagging and identification of multi-word expressions, and cast the task as a multi-task learning problem with deep recurrent neural networks. Our multi-task models perform significantly better than previous state of the art approaches on two scientific KBC datasets, particularly for long keyphrases.

AB - Keyphrase boundary classification (KBC) is the task of detecting keyphrases in scientific articles and labelling them with respect to predefined types. Although important in practice, this task is so far un-derexplored, partly due to the lack of labelled data. To overcome this, we explore several auxiliary tasks, including semantic super-sense tagging and identification of multi-word expressions, and cast the task as a multi-task learning problem with deep recurrent neural networks. Our multi-task models perform significantly better than previous state of the art approaches on two scientific KBC datasets, particularly for long keyphrases.

U2 - 10.18653/v1/p17-2054

DO - 10.18653/v1/p17-2054

M3 - Article in proceedings

SN - 978-1-945626-76-0

VL - 2

SP - 341

EP - 346

BT - Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics

A2 - Barzilay, Regina

A2 - Kan, Min-Yen

PB - Association for Computational Linguistics

T2 - Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

Y2 - 1 July 2017 through 1 July 2017

ER -

Multi-Task Learning of Keyphrase Boundary Classification

Abstract

Konference

Adgang til dokumentet

Fingeraftryk

Citationsformater