A novel neural network model for joint POS tagging and graph-based dependency parsing

Dat Quoc Nguyen; Mark Dras; Mark Johnson

doi:10.18653/v1/K17-3014

A novel neural network model for joint POS tagging and graph-based dependency parsing

Dat Quoc Nguyen, Mark Dras, Mark Johnson

School of Computing

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

26 Citations (Scopus)

Abstract

We present a novel neural network model that learns POS tagging and graph-based dependency parsing jointly. Our model uses bidirectional LSTMs to learn feature representations shared for both POS tagging and dependency parsing tasks, thus handling the feature-engineering problem. Our extensive experiments, on 19 languages from the Universal Dependencies project, show that our model outperforms the state-of-the-art neural network-based Stack-propagation model for joint POS tagging and transition-based dependency parsing, resulting in a new state of the art. Our code is open-source and available together with pre-trained models at: https://github.com/datquocnguyen/jPTDP

Original language	English
Title of host publication	Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies
Place of Publication	Stroudsburg PA
Publisher	Association for Computational Linguistics (ACL)
Pages	134-142
Number of pages	9
ISBN (Print)	9781945626708
DOIs	https://doi.org/10.18653/v1/K17-3014
Publication status	Published - 2017
Event	The SIGNLL Conference on Computational Natural Language Learning - Vancouver Duration: 3 Aug 2017 → 4 Aug 2017

Conference

Conference	The SIGNLL Conference on Computational Natural Language Learning
City	Vancouver
Period	3/08/17 → 4/08/17

Bibliographical note

Copyright the Publisher. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Keywords

neural network
POS taging
dependency parsing
bidirectional LSTM
universal dependencies
multilingual parsing

Access to Document

10.18653/v1/K17-3014

http://Publisher version (open access)Licence: CC BY

Cite this

@inproceedings{2cac9751cad24be380dff288887162f1,

title = "A novel neural network model for joint POS tagging and graph-based dependency parsing",

abstract = "We present a novel neural network model that learns POS tagging and graph-based dependency parsing jointly. Our model uses bidirectional LSTMs to learn feature representations shared for both POS tagging and dependency parsing tasks, thus handling the feature-engineering problem. Our extensive experiments, on 19 languages from the Universal Dependencies project, show that our model outperforms the state-of-the-art neural network-based Stack-propagation model for joint POS tagging and transition-based dependency parsing, resulting in a new state of the art. Our code is open-source and available together with pre-trained models at: https://github.com/datquocnguyen/jPTDP",

keywords = "neural network, POS taging, dependency parsing, bidirectional LSTM, universal dependencies, multilingual parsing",

author = "Nguyen, {Dat Quoc} and Mark Dras and Mark Johnson",

note = "Copyright the Publisher. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.; The SIGNLL Conference on Computational Natural Language Learning ; Conference date: 03-08-2017 Through 04-08-2017",

year = "2017",

doi = "10.18653/v1/K17-3014",

language = "English",

isbn = "9781945626708",

pages = "134--142",

booktitle = "Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies",

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

}

Nguyen, DQ, Dras, M & Johnson, M 2017, A novel neural network model for joint POS tagging and graph-based dependency parsing. in Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. Association for Computational Linguistics (ACL), Stroudsburg PA, pp. 134-142, The SIGNLL Conference on Computational Natural Language Learning, Vancouver, 3/08/17. https://doi.org/10.18653/v1/K17-3014

A novel neural network model for joint POS tagging and graph-based dependency parsing. / Nguyen, Dat Quoc; Dras, Mark ; Johnson, Mark.
Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. Stroudsburg PA: Association for Computational Linguistics (ACL), 2017. p. 134-142.

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

TY - GEN

T1 - A novel neural network model for joint POS tagging and graph-based dependency parsing

AU - Nguyen, Dat Quoc

AU - Dras, Mark

AU - Johnson, Mark

N1 - Copyright the Publisher. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

PY - 2017

Y1 - 2017

N2 - We present a novel neural network model that learns POS tagging and graph-based dependency parsing jointly. Our model uses bidirectional LSTMs to learn feature representations shared for both POS tagging and dependency parsing tasks, thus handling the feature-engineering problem. Our extensive experiments, on 19 languages from the Universal Dependencies project, show that our model outperforms the state-of-the-art neural network-based Stack-propagation model for joint POS tagging and transition-based dependency parsing, resulting in a new state of the art. Our code is open-source and available together with pre-trained models at: https://github.com/datquocnguyen/jPTDP

AB - We present a novel neural network model that learns POS tagging and graph-based dependency parsing jointly. Our model uses bidirectional LSTMs to learn feature representations shared for both POS tagging and dependency parsing tasks, thus handling the feature-engineering problem. Our extensive experiments, on 19 languages from the Universal Dependencies project, show that our model outperforms the state-of-the-art neural network-based Stack-propagation model for joint POS tagging and transition-based dependency parsing, resulting in a new state of the art. Our code is open-source and available together with pre-trained models at: https://github.com/datquocnguyen/jPTDP

KW - neural network

KW - POS taging

KW - dependency parsing

KW - bidirectional LSTM

KW - universal dependencies

KW - multilingual parsing

UR - http://www.scopus.com/inward/record.url?scp=85053871091&partnerID=8YFLogxK

U2 - 10.18653/v1/K17-3014

DO - 10.18653/v1/K17-3014

M3 - Conference proceeding contribution

SN - 9781945626708

SP - 134

EP - 142

BT - Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies

PB - Association for Computational Linguistics (ACL)

CY - Stroudsburg PA

T2 - The SIGNLL Conference on Computational Natural Language Learning

Y2 - 3 August 2017 through 4 August 2017

ER -

A novel neural network model for joint POS tagging and graph-based dependency parsing

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this