@inproceedings{schoene-etal-2024-models,
title = "All Models are Wrong, But Some are Deadly: Inconsistencies in Emotion Detection in Suicide-related Tweets",
author = "Schoene, Annika Marie and
Ramachandranpillai, Resmi and
Lazovich, Tomo and
Baeza-Yates, Ricardo A.",
editor = "Dementieva, Daryna and
Ignat, Oana and
Jin, Zhijing and
Mihalcea, Rada and
Piatti, Giorgio and
Tetreault, Joel and
Wilson, Steven and
Zhao, Jieyu",
booktitle = "Proceedings of the Third Workshop on NLP for Positive Impact",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.nlp4pi-1.9",
pages = "113--122",
abstract = "Recent work in psychology has shown that people who experience mental health challenges are more likely to express their thoughts, emotions, and feelings on social media than share it with a clinical professional. Distinguishing suicide-related content, such as suicide mentioned in a humorous context, from genuine expressions of suicidal ideation is essential to better understanding context and risk. In this paper, we give a first insight and analysis into the differences between emotion labels annotated by humans and labels predicted by three fine-tuned language models (LMs) for suicide-related content. We find that (i) there is little agreement between LMs and humans for emotion labels of suicide-related Tweets and (ii) individual LMs predict similar emotion labels for all suicide-related categories. Our findings lead us to question the credibility and usefulness of such methods in high-risk scenarios such as suicide ideation detection.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="schoene-etal-2024-models">
<titleInfo>
<title>All Models are Wrong, But Some are Deadly: Inconsistencies in Emotion Detection in Suicide-related Tweets</title>
</titleInfo>
<name type="personal">
<namePart type="given">Annika</namePart>
<namePart type="given">Marie</namePart>
<namePart type="family">Schoene</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Resmi</namePart>
<namePart type="family">Ramachandranpillai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tomo</namePart>
<namePart type="family">Lazovich</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ricardo</namePart>
<namePart type="given">A</namePart>
<namePart type="family">Baeza-Yates</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Third Workshop on NLP for Positive Impact</title>
</titleInfo>
<name type="personal">
<namePart type="given">Daryna</namePart>
<namePart type="family">Dementieva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Oana</namePart>
<namePart type="family">Ignat</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhijing</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rada</namePart>
<namePart type="family">Mihalcea</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Giorgio</namePart>
<namePart type="family">Piatti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joel</namePart>
<namePart type="family">Tetreault</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Wilson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jieyu</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Recent work in psychology has shown that people who experience mental health challenges are more likely to express their thoughts, emotions, and feelings on social media than share it with a clinical professional. Distinguishing suicide-related content, such as suicide mentioned in a humorous context, from genuine expressions of suicidal ideation is essential to better understanding context and risk. In this paper, we give a first insight and analysis into the differences between emotion labels annotated by humans and labels predicted by three fine-tuned language models (LMs) for suicide-related content. We find that (i) there is little agreement between LMs and humans for emotion labels of suicide-related Tweets and (ii) individual LMs predict similar emotion labels for all suicide-related categories. Our findings lead us to question the credibility and usefulness of such methods in high-risk scenarios such as suicide ideation detection.</abstract>
<identifier type="citekey">schoene-etal-2024-models</identifier>
<location>
<url>https://aclanthology.org/2024.nlp4pi-1.9</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>113</start>
<end>122</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T All Models are Wrong, But Some are Deadly: Inconsistencies in Emotion Detection in Suicide-related Tweets
%A Schoene, Annika Marie
%A Ramachandranpillai, Resmi
%A Lazovich, Tomo
%A Baeza-Yates, Ricardo A.
%Y Dementieva, Daryna
%Y Ignat, Oana
%Y Jin, Zhijing
%Y Mihalcea, Rada
%Y Piatti, Giorgio
%Y Tetreault, Joel
%Y Wilson, Steven
%Y Zhao, Jieyu
%S Proceedings of the Third Workshop on NLP for Positive Impact
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F schoene-etal-2024-models
%X Recent work in psychology has shown that people who experience mental health challenges are more likely to express their thoughts, emotions, and feelings on social media than share it with a clinical professional. Distinguishing suicide-related content, such as suicide mentioned in a humorous context, from genuine expressions of suicidal ideation is essential to better understanding context and risk. In this paper, we give a first insight and analysis into the differences between emotion labels annotated by humans and labels predicted by three fine-tuned language models (LMs) for suicide-related content. We find that (i) there is little agreement between LMs and humans for emotion labels of suicide-related Tweets and (ii) individual LMs predict similar emotion labels for all suicide-related categories. Our findings lead us to question the credibility and usefulness of such methods in high-risk scenarios such as suicide ideation detection.
%U https://aclanthology.org/2024.nlp4pi-1.9
%P 113-122
Markdown (Informal)
[All Models are Wrong, But Some are Deadly: Inconsistencies in Emotion Detection in Suicide-related Tweets](https://aclanthology.org/2024.nlp4pi-1.9) (Schoene et al., NLP4PI 2024)
ACL