Domain Information Enhanced Dependency Parser

Nan Yu; Zonglin Liu; Ranran Zhen; Tao Liu; Meishan Zhang; Guohong Fu

doi:10.1007/978-3-030-32236-6_73

Domain Information Enhanced Dependency Parser

Nan Yu
, Zonglin Liu
, Ranran Zhen
, Tao Liu
, Meishan Zhang
, Guohong Fu^*

^*Corresponding author for this work

Soochow University
Heilongjiang University
Tianjin University

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

2 Scopus citations

Abstract

Dependency parsing has been an important task in the natural language processing (NLP) community. Supervised methods have achieved great success these years. However, these models can suffer significant performance loss when test domain differs from the training domain. In this paper, we adopt the Bi-Affine parser as our baseline. To explore domain-specific information and domain-independent information for cross-domain dependency parsing, we apply an ensemble-style self-training and adversarial learning, respectively. We finally combine the two strategies to enhance our baseline model and our final system was ranked the first of at NLPCC2019 shared task on cross-domain dependency parsing.

Original language	English
Title of host publication	Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings
Editors	Jie Tang, Min-Yen Kan, Dongyan Zhao, Sujian Li, Hongying Zan
Publisher	Springer
Pages	801-810
Number of pages	10
ISBN (Print)	9783030322359
DOIs	https://doi.org/10.1007/978-3-030-32236-6_73
State	Published - 2019
Externally published	Yes
Event	8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019 - Dunhuang, China Duration: 9 Oct 2019 → 14 Oct 2019

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11839 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019
Country/Territory	China
City	Dunhuang
Period	9/10/19 → 14/10/19

Keywords

Adversarial learning
Cross-domain
Dependency parsing
Ensemble
Self-training

Access to Document

10.1007/978-3-030-32236-6_73

Cite this

Yu, N., Liu, Z., Zhen, R., Liu, T., Zhang, M., & Fu, G. (2019). Domain Information Enhanced Dependency Parser. In J. Tang, M.-Y. Kan, D. Zhao, S. Li, & H. Zan (Eds.), Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings (pp. 801-810). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11839 LNAI). Springer. https://doi.org/10.1007/978-3-030-32236-6_73

Yu, Nan ; Liu, Zonglin ; Zhen, Ranran et al. / Domain Information Enhanced Dependency Parser. Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings. editor / Jie Tang ; Min-Yen Kan ; Dongyan Zhao ; Sujian Li ; Hongying Zan. Springer, 2019. pp. 801-810 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{607c3494706f44afad53318daa8d1e91,

title = "Domain Information Enhanced Dependency Parser",

abstract = "Dependency parsing has been an important task in the natural language processing (NLP) community. Supervised methods have achieved great success these years. However, these models can suffer significant performance loss when test domain differs from the training domain. In this paper, we adopt the Bi-Affine parser as our baseline. To explore domain-specific information and domain-independent information for cross-domain dependency parsing, we apply an ensemble-style self-training and adversarial learning, respectively. We finally combine the two strategies to enhance our baseline model and our final system was ranked the first of at NLPCC2019 shared task on cross-domain dependency parsing.",

keywords = "Adversarial learning, Cross-domain, Dependency parsing, Ensemble, Self-training",

author = "Nan Yu and Zonglin Liu and Ranran Zhen and Tao Liu and Meishan Zhang and Guohong Fu",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019 ; Conference date: 09-10-2019 Through 14-10-2019",

year = "2019",

doi = "10.1007/978-3-030-32236-6\_73",

language = "英语",

isbn = "9783030322359",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "801--810",

editor = "Jie Tang and Min-Yen Kan and Dongyan Zhao and Sujian Li and Hongying Zan",

booktitle = "Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings",

address = "德国",

}

Yu, N, Liu, Z, Zhen, R, Liu, T, Zhang, M & Fu, G 2019, Domain Information Enhanced Dependency Parser. in J Tang, M-Y Kan, D Zhao, S Li & H Zan (eds), Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11839 LNAI, Springer, pp. 801-810, 8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019, Dunhuang, China, 9/10/19. https://doi.org/10.1007/978-3-030-32236-6_73

Domain Information Enhanced Dependency Parser. / Yu, Nan; Liu, Zonglin; Zhen, Ranran et al.
Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings. ed. / Jie Tang; Min-Yen Kan; Dongyan Zhao; Sujian Li; Hongying Zan. Springer, 2019. p. 801-810 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11839 LNAI).

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

TY - GEN

T1 - Domain Information Enhanced Dependency Parser

AU - Yu, Nan

AU - Liu, Zonglin

AU - Zhen, Ranran

AU - Liu, Tao

AU - Zhang, Meishan

AU - Fu, Guohong

PY - 2019

Y1 - 2019

N2 - Dependency parsing has been an important task in the natural language processing (NLP) community. Supervised methods have achieved great success these years. However, these models can suffer significant performance loss when test domain differs from the training domain. In this paper, we adopt the Bi-Affine parser as our baseline. To explore domain-specific information and domain-independent information for cross-domain dependency parsing, we apply an ensemble-style self-training and adversarial learning, respectively. We finally combine the two strategies to enhance our baseline model and our final system was ranked the first of at NLPCC2019 shared task on cross-domain dependency parsing.

AB - Dependency parsing has been an important task in the natural language processing (NLP) community. Supervised methods have achieved great success these years. However, these models can suffer significant performance loss when test domain differs from the training domain. In this paper, we adopt the Bi-Affine parser as our baseline. To explore domain-specific information and domain-independent information for cross-domain dependency parsing, we apply an ensemble-style self-training and adversarial learning, respectively. We finally combine the two strategies to enhance our baseline model and our final system was ranked the first of at NLPCC2019 shared task on cross-domain dependency parsing.

KW - Adversarial learning

KW - Cross-domain

KW - Dependency parsing

KW - Ensemble

KW - Self-training

UR - https://www.scopus.com/pages/publications/85075817048

U2 - 10.1007/978-3-030-32236-6_73

DO - 10.1007/978-3-030-32236-6_73

M3 - 会议稿件

AN - SCOPUS:85075817048

SN - 9783030322359

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 801

EP - 810

BT - Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings

A2 - Tang, Jie

A2 - Kan, Min-Yen

A2 - Zhao, Dongyan

A2 - Li, Sujian

A2 - Zan, Hongying

PB - Springer

T2 - 8th CCF International Conference on Natural Language Processing and Chinese Computing, NLPCC 2019

Y2 - 9 October 2019 through 14 October 2019

ER -

Yu N, Liu Z, Zhen R, Liu T, Zhang M, Fu G. Domain Information Enhanced Dependency Parser. In Tang J, Kan MY, Zhao D, Li S, Zan H, editors, Natural Language Processing and Chinese Computing - 8th CCF International Conference, NLPCC 2019, Proceedings. Springer. 2019. p. 801-810. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-32236-6_73

Domain Information Enhanced Dependency Parser

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this