@inproceedings{weld-etal-2021-leveraging,
title = "Leveraging Community and Author Context to Explain the Performance and Bias of Text-Based Deception Detection Models",
author = "Weld, Galen and
Ayton, Ellyn and
Althoff, Tim and
Glenski, Maria",
editor = "Feldman, Anna and
Da San Martino, Giovanni and
Leberknight, Chris and
Nakov, Preslav",
booktitle = "Proceedings of the Fourth Workshop on NLP for Internet Freedom: Censorship, Disinformation, and Propaganda",
month = jun,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.nlp4if-1.5",
doi = "10.18653/v1/2021.nlp4if-1.5",
pages = "29--38",
abstract = "Deceptive news posts shared in online communities can be detected with NLP models, and much recent research has focused on the development of such models. In this work, we use characteristics of online communities and authors {---} the context of how and where content is posted {---} to explain the performance of a neural network deception detection model and identify sub-populations who are disproportionately affected by model accuracy or failure. We examine who is posting the content, and where the content is posted to. We find that while author characteristics are better predictors of deceptive content than community characteristics, both characteristics are strongly correlated with model performance. Traditional performance metrics such as F1 score may fail to capture poor model performance on isolated sub-populations such as specific authors, and as such, more nuanced evaluation of deception detection models is critical.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="weld-etal-2021-leveraging">
<titleInfo>
<title>Leveraging Community and Author Context to Explain the Performance and Bias of Text-Based Deception Detection Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Galen</namePart>
<namePart type="family">Weld</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ellyn</namePart>
<namePart type="family">Ayton</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tim</namePart>
<namePart type="family">Althoff</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Glenski</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-06</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fourth Workshop on NLP for Internet Freedom: Censorship, Disinformation, and Propaganda</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Feldman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Giovanni</namePart>
<namePart type="family">Da San Martino</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chris</namePart>
<namePart type="family">Leberknight</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Preslav</namePart>
<namePart type="family">Nakov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Deceptive news posts shared in online communities can be detected with NLP models, and much recent research has focused on the development of such models. In this work, we use characteristics of online communities and authors — the context of how and where content is posted — to explain the performance of a neural network deception detection model and identify sub-populations who are disproportionately affected by model accuracy or failure. We examine who is posting the content, and where the content is posted to. We find that while author characteristics are better predictors of deceptive content than community characteristics, both characteristics are strongly correlated with model performance. Traditional performance metrics such as F1 score may fail to capture poor model performance on isolated sub-populations such as specific authors, and as such, more nuanced evaluation of deception detection models is critical.</abstract>
<identifier type="citekey">weld-etal-2021-leveraging</identifier>
<identifier type="doi">10.18653/v1/2021.nlp4if-1.5</identifier>
<location>
<url>https://aclanthology.org/2021.nlp4if-1.5</url>
</location>
<part>
<date>2021-06</date>
<extent unit="page">
<start>29</start>
<end>38</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Leveraging Community and Author Context to Explain the Performance and Bias of Text-Based Deception Detection Models
%A Weld, Galen
%A Ayton, Ellyn
%A Althoff, Tim
%A Glenski, Maria
%Y Feldman, Anna
%Y Da San Martino, Giovanni
%Y Leberknight, Chris
%Y Nakov, Preslav
%S Proceedings of the Fourth Workshop on NLP for Internet Freedom: Censorship, Disinformation, and Propaganda
%D 2021
%8 June
%I Association for Computational Linguistics
%C Online
%F weld-etal-2021-leveraging
%X Deceptive news posts shared in online communities can be detected with NLP models, and much recent research has focused on the development of such models. In this work, we use characteristics of online communities and authors — the context of how and where content is posted — to explain the performance of a neural network deception detection model and identify sub-populations who are disproportionately affected by model accuracy or failure. We examine who is posting the content, and where the content is posted to. We find that while author characteristics are better predictors of deceptive content than community characteristics, both characteristics are strongly correlated with model performance. Traditional performance metrics such as F1 score may fail to capture poor model performance on isolated sub-populations such as specific authors, and as such, more nuanced evaluation of deception detection models is critical.
%R 10.18653/v1/2021.nlp4if-1.5
%U https://aclanthology.org/2021.nlp4if-1.5
%U https://doi.org/10.18653/v1/2021.nlp4if-1.5
%P 29-38
Markdown (Informal)
[Leveraging Community and Author Context to Explain the Performance and Bias of Text-Based Deception Detection Models](https://aclanthology.org/2021.nlp4if-1.5) (Weld et al., NLP4IF 2021)
ACL