Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning

Jun Li; Zhihan Yan; Zehang Lin; Xingyun Liu; Hong Va Leong; Nancy Xiaonan Yu; Qing Li

doi:10.1007/978-3-030-85896-4_12

Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning

Jun Li
, Zhihan Yan
, Zehang Lin
, Xingyun Liu
, Hong Va Leong
, Nancy Xiaonan Yu
, Qing Li

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

5 Citations (Scopus)

Abstract

Suicide ideation detection on social media is a challenging problem due to its implicitness. In this paper, we present an approach to detect suicide ideation on social media based on a BERT-LSTM model with Adversarial and Multi-task learning (BLAM). More specifically, BLAM combines BERT model with Bi-LSTM model to extract deeper and richer features. Furthermore, emotion classification is utilized as an auxiliary task to perform multi-task learning, which enriches the extracted features with emotion information that enhances the identification of suicide. In addition, BLAM generates adversarial noise by adversarial learning improving the generalization ability of the model. Extensive experiments conducted on our collected Suicide Ideation Detection (SID) dataset demonstrate the competitive superiority of BLAM compared with the state-of-the-art methods.

Original language	English
Title of host publication	Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings
Editors	Leong Hou U, Marc Spaniol, Yasushi Sakurai, Junying Chen
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	140-145
Number of pages	6
ISBN (Print)	9783030858957
DOIs	https://doi.org/10.1007/978-3-030-85896-4_12
Publication status	Published - 2021
Event	5th International Joint Conference on Asia-Paciﬁc Web and Web-Age Information Management, APWeb-WAIM 2021 - Guangzhou, China Duration: 23 Aug 2021 → 25 Aug 2021

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12858 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	5th International Joint Conference on Asia-Paciﬁc Web and Web-Age Information Management, APWeb-WAIM 2021
Country/Territory	China
City	Guangzhou
Period	23/08/21 → 25/08/21

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Adversarial learning
Multi-task learning
Suicide ideation detection

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

More information

10.1007/978-3-030-85896-4_12

Cite this

Li, J., Yan, Z., Lin, Z., Liu, X., Leong, H. V., Yu, N. X., & Li, Q. (2021). Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning. In L. H. U, M. Spaniol, Y. Sakurai, & J. Chen (Eds.), Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings (pp. 140-145). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12858 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-85896-4_12

Li, Jun ; Yan, Zhihan ; Lin, Zehang et al. / Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning. Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings. editor / Leong Hou U ; Marc Spaniol ; Yasushi Sakurai ; Junying Chen. Springer Science and Business Media Deutschland GmbH, 2021. pp. 140-145 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{ca67ed3875dc457cb1203f4f920fa7a0,

title = "Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning",

abstract = "Suicide ideation detection on social media is a challenging problem due to its implicitness. In this paper, we present an approach to detect suicide ideation on social media based on a BERT-LSTM model with Adversarial and Multi-task learning (BLAM). More specifically, BLAM combines BERT model with Bi-LSTM model to extract deeper and richer features. Furthermore, emotion classification is utilized as an auxiliary task to perform multi-task learning, which enriches the extracted features with emotion information that enhances the identification of suicide. In addition, BLAM generates adversarial noise by adversarial learning improving the generalization ability of the model. Extensive experiments conducted on our collected Suicide Ideation Detection (SID) dataset demonstrate the competitive superiority of BLAM compared with the state-of-the-art methods.",

keywords = "Adversarial learning, Multi-task learning, Suicide ideation detection",

author = "Jun Li and Zhihan Yan and Zehang Lin and Xingyun Liu and Leong, \{Hong Va\} and Yu, \{Nancy Xiaonan\} and Qing Li",

note = "Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.; 5th International Joint Conference on Asia-Paciﬁc Web and Web-Age Information Management, APWeb-WAIM 2021 ; Conference date: 23-08-2021 Through 25-08-2021",

year = "2021",

doi = "10.1007/978-3-030-85896-4\_12",

language = "English",

isbn = "9783030858957",

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

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "140--145",

editor = "U, \{Leong Hou\} and Marc Spaniol and Yasushi Sakurai and Junying Chen",

booktitle = "Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings",

address = "Germany",

}

Li, J, Yan, Z, Lin, Z, Liu, X, Leong, HV, Yu, NX & Li, Q 2021, Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning. in LH U, M Spaniol, Y Sakurai & J Chen (eds), Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12858 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 140-145, 5th International Joint Conference on Asia-Paciﬁc Web and Web-Age Information Management, APWeb-WAIM 2021, Guangzhou, China, 23/08/21. https://doi.org/10.1007/978-3-030-85896-4_12

Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning. / Li, Jun; Yan, Zhihan; Lin, Zehang et al.
Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings. ed. / Leong Hou U; Marc Spaniol; Yasushi Sakurai; Junying Chen. Springer Science and Business Media Deutschland GmbH, 2021. p. 140-145 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12858 LNCS).

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning

AU - Li, Jun

AU - Yan, Zhihan

AU - Lin, Zehang

AU - Liu, Xingyun

AU - Leong, Hong Va

AU - Yu, Nancy Xiaonan

AU - Li, Qing

PY - 2021

Y1 - 2021

N2 - Suicide ideation detection on social media is a challenging problem due to its implicitness. In this paper, we present an approach to detect suicide ideation on social media based on a BERT-LSTM model with Adversarial and Multi-task learning (BLAM). More specifically, BLAM combines BERT model with Bi-LSTM model to extract deeper and richer features. Furthermore, emotion classification is utilized as an auxiliary task to perform multi-task learning, which enriches the extracted features with emotion information that enhances the identification of suicide. In addition, BLAM generates adversarial noise by adversarial learning improving the generalization ability of the model. Extensive experiments conducted on our collected Suicide Ideation Detection (SID) dataset demonstrate the competitive superiority of BLAM compared with the state-of-the-art methods.

AB - Suicide ideation detection on social media is a challenging problem due to its implicitness. In this paper, we present an approach to detect suicide ideation on social media based on a BERT-LSTM model with Adversarial and Multi-task learning (BLAM). More specifically, BLAM combines BERT model with Bi-LSTM model to extract deeper and richer features. Furthermore, emotion classification is utilized as an auxiliary task to perform multi-task learning, which enriches the extracted features with emotion information that enhances the identification of suicide. In addition, BLAM generates adversarial noise by adversarial learning improving the generalization ability of the model. Extensive experiments conducted on our collected Suicide Ideation Detection (SID) dataset demonstrate the competitive superiority of BLAM compared with the state-of-the-art methods.

KW - Adversarial learning

KW - Multi-task learning

KW - Suicide ideation detection

UR - https://www.scopus.com/pages/publications/85115162508

U2 - 10.1007/978-3-030-85896-4_12

DO - 10.1007/978-3-030-85896-4_12

M3 - Conference article published in proceeding or book

AN - SCOPUS:85115162508

SN - 9783030858957

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

SP - 140

EP - 145

BT - Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings

A2 - U, Leong Hou

A2 - Spaniol, Marc

A2 - Sakurai, Yasushi

A2 - Chen, Junying

PB - Springer Science and Business Media Deutschland GmbH

T2 - 5th International Joint Conference on Asia-Paciﬁc Web and Web-Age Information Management, APWeb-WAIM 2021

Y2 - 23 August 2021 through 25 August 2021

ER -

Li J, Yan Z, Lin Z, Liu X, Leong HV, Yu NX et al. Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning. In U LH, Spaniol M, Sakurai Y, Chen J, editors, Web and Big Data - 5th International Joint Conference, APWeb-WAIM 2021, Proceedings. Springer Science and Business Media Deutschland GmbH. 2021. p. 140-145. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-85896-4_12

Suicide Ideation Detection on Social Media During COVID-19 via Adversarial and Multi-task Learning

Abstract

Publication series

Conference

UN SDGs

Keywords

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this