Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task

Avi Gamoran; Yonatan Kaplan; Ram Isaac Orr; Almog Simchon; Michael Gilead

Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task

Avi Gamoran, Yonatan Kaplan, Ram Isaac Orr, Almog Simchon, Michael Gilead

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

Abstract

This paper describes our approach to the CLPsych 2021 Shared Task, in which we aimed to predict suicide attempts based on Twitter feed data. We addressed this challenge by emphasizing reliance on prior domain knowledge. We engineered novel theorydriven features, and integrated prior knowledge with empirical evidence in a principled manner using Bayesian modeling. While this theory-guided approach increases bias and lowers accuracy on the training set, it was successful in preventing over-fitting. The models provided reasonable classification accuracy on unseen test data (0:68 AUC 0:84). Our approach may be particularly useful in prediction tasks trained on a relatively small data set.

Original language	English
Title of host publication	Computational Linguistics and Clinical Psychology
Subtitle of host publication	Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021
Editors	Nazli Goharian, Philip Resnik, Andrew Yates, Molly Ireland, Kate Niederhoffer, Rebecca Resnik
Publisher	Association for Computational Linguistics (ACL)
Pages	103-109
Number of pages	7
ISBN (Electronic)	9781954085411
State	Published - 1 Jan 2021
Event	7th Workshop on Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Virtual, Online Duration: 11 Jun 2021 → …

Publication series

Name	Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021

Conference

Conference	7th Workshop on Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021
City	Virtual, Online
Period	11/06/21 → …

UN SDGs

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

SDG 3 Good Health and Well-being

All Science Journal Classification (ASJC) codes

Language and Linguistics
Computer Networks and Communications
Speech and Hearing

Cite this

Gamoran, A., Kaplan, Y., Orr, R. I., Simchon, A., & Gilead, M. (2021). Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task. In N. Goharian, P. Resnik, A. Yates, M. Ireland, K. Niederhoffer, & R. Resnik (Eds.), Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021 (pp. 103-109). (Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021). Association for Computational Linguistics (ACL).

Gamoran, Avi ; Kaplan, Yonatan ; Orr, Ram Isaac et al. / Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task. Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021. editor / Nazli Goharian ; Philip Resnik ; Andrew Yates ; Molly Ireland ; Kate Niederhoffer ; Rebecca Resnik. Association for Computational Linguistics (ACL), 2021. pp. 103-109 (Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021).

@inproceedings{5630f7bbb0a74b89b587473558245eae,

title = "Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task",

abstract = "This paper describes our approach to the CLPsych 2021 Shared Task, in which we aimed to predict suicide attempts based on Twitter feed data. We addressed this challenge by emphasizing reliance on prior domain knowledge. We engineered novel theorydriven features, and integrated prior knowledge with empirical evidence in a principled manner using Bayesian modeling. While this theory-guided approach increases bias and lowers accuracy on the training set, it was successful in preventing over-fitting. The models provided reasonable classification accuracy on unseen test data (0:68 AUC 0:84). Our approach may be particularly useful in prediction tasks trained on a relatively small data set.",

author = "Avi Gamoran and Yonatan Kaplan and Orr, \{Ram Isaac\} and Almog Simchon and Michael Gilead",

note = "Publisher Copyright: {\textcopyright}2021 Association for Computational Linguistics.; 7th Workshop on Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 ; Conference date: 11-06-2021",

year = "2021",

month = jan,

day = "1",

language = "الإنجليزيّة",

series = "Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021",

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

pages = "103--109",

editor = "Nazli Goharian and Philip Resnik and Andrew Yates and Molly Ireland and Kate Niederhoffer and Rebecca Resnik",

booktitle = "Computational Linguistics and Clinical Psychology",

address = "الولايات المتّحدة",

}

Gamoran, A, Kaplan, Y, Orr, RI, Simchon, A & Gilead, M 2021, Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task. in N Goharian, P Resnik, A Yates, M Ireland, K Niederhoffer & R Resnik (eds), Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021. Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021, Association for Computational Linguistics (ACL), pp. 103-109, 7th Workshop on Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021, Virtual, Online, 11/06/21.

Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task. / Gamoran, Avi; Kaplan, Yonatan; Orr, Ram Isaac et al.
Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021. ed. / Nazli Goharian; Philip Resnik; Andrew Yates; Molly Ireland; Kate Niederhoffer; Rebecca Resnik. Association for Computational Linguistics (ACL), 2021. p. 103-109 (Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021).

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

TY - GEN

T1 - Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task

AU - Gamoran, Avi

AU - Kaplan, Yonatan

AU - Orr, Ram Isaac

AU - Simchon, Almog

AU - Gilead, Michael

PY - 2021/1/1

Y1 - 2021/1/1

N2 - This paper describes our approach to the CLPsych 2021 Shared Task, in which we aimed to predict suicide attempts based on Twitter feed data. We addressed this challenge by emphasizing reliance on prior domain knowledge. We engineered novel theorydriven features, and integrated prior knowledge with empirical evidence in a principled manner using Bayesian modeling. While this theory-guided approach increases bias and lowers accuracy on the training set, it was successful in preventing over-fitting. The models provided reasonable classification accuracy on unseen test data (0:68 AUC 0:84). Our approach may be particularly useful in prediction tasks trained on a relatively small data set.

AB - This paper describes our approach to the CLPsych 2021 Shared Task, in which we aimed to predict suicide attempts based on Twitter feed data. We addressed this challenge by emphasizing reliance on prior domain knowledge. We engineered novel theorydriven features, and integrated prior knowledge with empirical evidence in a principled manner using Bayesian modeling. While this theory-guided approach increases bias and lowers accuracy on the training set, it was successful in preventing over-fitting. The models provided reasonable classification accuracy on unseen test data (0:68 AUC 0:84). Our approach may be particularly useful in prediction tasks trained on a relatively small data set.

UR - http://www.scopus.com/inward/record.url?scp=85112801937&partnerID=8YFLogxK

M3 - منشور من مؤتمر

T3 - Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021

SP - 103

EP - 109

BT - Computational Linguistics and Clinical Psychology

A2 - Goharian, Nazli

A2 - Resnik, Philip

A2 - Yates, Andrew

A2 - Ireland, Molly

A2 - Niederhoffer, Kate

A2 - Resnik, Rebecca

PB - Association for Computational Linguistics (ACL)

T2 - 7th Workshop on Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021

Y2 - 11 June 2021

ER -

Gamoran A, Kaplan Y, Orr RI, Simchon A , Gilead M. Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task. In Goharian N, Resnik P, Yates A, Ireland M, Niederhoffer K, Resnik R, editors, Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021. Association for Computational Linguistics (ACL). 2021. p. 103-109. (Computational Linguistics and Clinical Psychology: Improving Access, CLPsych 2021 - Proceedings of the 7th Workshop, in conjunction with NAACL 2021).

Using Psychologically-Informed Priors for Suicide Prediction in the CLPsych 2021 Shared Task

Abstract

Publication series

Conference

UN SDGs

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this