@inproceedings{ambalavanan-etal-2019-using,
title = "Using Contextual Representations for Suicide Risk Assessment from {I}nternet Forums",
author = "Ambalavanan, Ashwin Karthik and
Jagtap, Pranjali Dileep and
Adhya, Soumya and
Devarakonda, Murthy",
editor = "Niederhoffer, Kate and
Hollingshead, Kristy and
Resnik, Philip and
Resnik, Rebecca and
Loveys, Kate",
booktitle = "Proceedings of the Sixth Workshop on Computational Linguistics and Clinical Psychology",
month = jun,
year = "2019",
address = "Minneapolis, Minnesota",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/W19-3022/",
doi = "10.18653/v1/W19-3022",
pages = "172--176",
abstract = "Social media posts may yield clues to the subject{'}s (usually, the writer{'}s) suicide risk and intent, which can be used for timely intervention. This research, motivated by the CLPsych 2019 shared task, developed neural network-based methods for analyzing posts in one or more Reddit forums to assess the subject{'}s suicide risk. One of the technical challenges this task poses is the large amount of text from multiple posts of a single user. Our neural network models use the advanced multi-headed Attention-based autoencoder architecture, called Bidirectional Encoder Representations from Transformers (BERT). Our system achieved the 2nd best performance of 0.477 macro averaged F measure on Task A of the challenge. Among the three different alternatives we developed for the challenge, the single BERT model that processed all of a user{'}s posts performed the best on all three Tasks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ambalavanan-etal-2019-using">
<titleInfo>
<title>Using Contextual Representations for Suicide Risk Assessment from Internet Forums</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ashwin</namePart>
<namePart type="given">Karthik</namePart>
<namePart type="family">Ambalavanan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pranjali</namePart>
<namePart type="given">Dileep</namePart>
<namePart type="family">Jagtap</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Soumya</namePart>
<namePart type="family">Adhya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Murthy</namePart>
<namePart type="family">Devarakonda</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2019-06</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Sixth Workshop on Computational Linguistics and Clinical Psychology</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kate</namePart>
<namePart type="family">Niederhoffer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kristy</namePart>
<namePart type="family">Hollingshead</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philip</namePart>
<namePart type="family">Resnik</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rebecca</namePart>
<namePart type="family">Resnik</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kate</namePart>
<namePart type="family">Loveys</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Minneapolis, Minnesota</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Social media posts may yield clues to the subject’s (usually, the writer’s) suicide risk and intent, which can be used for timely intervention. This research, motivated by the CLPsych 2019 shared task, developed neural network-based methods for analyzing posts in one or more Reddit forums to assess the subject’s suicide risk. One of the technical challenges this task poses is the large amount of text from multiple posts of a single user. Our neural network models use the advanced multi-headed Attention-based autoencoder architecture, called Bidirectional Encoder Representations from Transformers (BERT). Our system achieved the 2nd best performance of 0.477 macro averaged F measure on Task A of the challenge. Among the three different alternatives we developed for the challenge, the single BERT model that processed all of a user’s posts performed the best on all three Tasks.</abstract>
<identifier type="citekey">ambalavanan-etal-2019-using</identifier>
<identifier type="doi">10.18653/v1/W19-3022</identifier>
<location>
<url>https://aclanthology.org/W19-3022/</url>
</location>
<part>
<date>2019-06</date>
<extent unit="page">
<start>172</start>
<end>176</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Using Contextual Representations for Suicide Risk Assessment from Internet Forums
%A Ambalavanan, Ashwin Karthik
%A Jagtap, Pranjali Dileep
%A Adhya, Soumya
%A Devarakonda, Murthy
%Y Niederhoffer, Kate
%Y Hollingshead, Kristy
%Y Resnik, Philip
%Y Resnik, Rebecca
%Y Loveys, Kate
%S Proceedings of the Sixth Workshop on Computational Linguistics and Clinical Psychology
%D 2019
%8 June
%I Association for Computational Linguistics
%C Minneapolis, Minnesota
%F ambalavanan-etal-2019-using
%X Social media posts may yield clues to the subject’s (usually, the writer’s) suicide risk and intent, which can be used for timely intervention. This research, motivated by the CLPsych 2019 shared task, developed neural network-based methods for analyzing posts in one or more Reddit forums to assess the subject’s suicide risk. One of the technical challenges this task poses is the large amount of text from multiple posts of a single user. Our neural network models use the advanced multi-headed Attention-based autoencoder architecture, called Bidirectional Encoder Representations from Transformers (BERT). Our system achieved the 2nd best performance of 0.477 macro averaged F measure on Task A of the challenge. Among the three different alternatives we developed for the challenge, the single BERT model that processed all of a user’s posts performed the best on all three Tasks.
%R 10.18653/v1/W19-3022
%U https://aclanthology.org/W19-3022/
%U https://doi.org/10.18653/v1/W19-3022
%P 172-176
Markdown (Informal)
[Using Contextual Representations for Suicide Risk Assessment from Internet Forums](https://aclanthology.org/W19-3022/) (Ambalavanan et al., CLPsych 2019)
ACL