Introducing the Gab Hate Corpus: defining and applying hate-based rhetoric to social media posts at scale

Kennedy, Brendan; Atari, Mohammad; Davani, Aida Mostafazadeh; Yeh, Leigh; Omrani, Ali S.; Kim, Yehsong; Coombs, Kris; Havaldar, Shreya; Portillo-Wightman, Gwenyth; Gonzalez, Elaine; Hoover, Joe; Azatian, Aida; Hussain, Alyzeh; Lara, Austin; Cardenas, Gabriel; Omary, Adam; Park, Christina; Wang, Xin; Wijaya, Clarisa; Yong, Zhang; Meyerowitz, Beth E.; Dehghani, Morteza

doi:10.1007/s10579-021-09569-x

Cited by 24 publications

(19 citation statements)

References 63 publications

(72 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To simulate human feedback for the instance explanation, we leverage ground truth rationales for SST (Carton et al, 2020) as human feedback. For hate speech detection, we use STF (de Gibert et al, 2018) as the ID dataset, and HatEval (Barbieri et al, 2020), Gab Hate Corpus (GHC) (Kennedy et al, 2018) and Latent Hatred (ElSherief et al, 2021) for OOD datasets. To simulate human feedback for the task explanations, we leverage group identifiers (e.g., black, muslims) (Kennedy et al, 2020b) be discarded for determining whether the instance is hate or not.…”

Section: Methodsmentioning

confidence: 99%

XMD: An End-to-End Framework for Interactive Explanation-Based Debugging of NLP Models

Lee¹,

Akshen²,

Joshi³

et al. 2022

Preprint

View full text Add to dashboard Cite

NLP models are susceptible to learning spurious biases (i.e., bugs) that work on some datasets but do not properly reflect the underlying task. Explanation-based model debugging aims to resolve spurious biases by showing human users explanations of model behavior, asking users to give feedback on the behavior, then using the feedback to update the model. While existing model debugging methods have shown promise, their prototypelevel implementations provide limited practical utility. Thus, we propose XMD : the first open-source, end-to-end framework for explanation-based model debugging. Given task-or instance-level explanations, users can flexibly provide various forms of feedback via an intuitive, web-based UI. After receiving user feedback, XMD automatically updates the model in real time, by regularizing the model so that its explanations align with the user feedback. The new model can then be easily deployed into real-world applications via Hugging Face. Using XMD , we can improve the model's OOD performance on text classification tasks by up to 18%. 1 * Both authors contributed equally.

show abstract

Section: Methodsmentioning

confidence: 99%

XMD: An End-to-End Framework for Interactive Explanation-Based Debugging of NLP Models

Lee¹,

Akshen²,

Joshi³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Our account relies on a sociological perspective that accentuates a social and cultural construction of the hateful rhetoric. As Kennedy et al [ 28 ] observed, “hate speech is fundamentally embedded within the existing cultural and social context in which it occurs”. Consequently, the weights assigned to each category of hatred and their temporal shifts that we focus on here might be largely context specific, reflecting political and cultural dynamics that are central to South Korea.…”

Section: Theoretical Backgroundmentioning

confidence: 99%

The hatred of all against all? Evidence from online community platforms in South Korea

Koo,

Suh,

Chung

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

Background Over several years of recent efforts to make sense and detect online hate speech, we still know relatively little about how hateful expressions enter online platforms and whether there are patterns and features characterizing the corpus of hateful speech. Objective In this research, we introduce a new conceptual framework suitable for better capturing the overall scope and dynamics of the current forms of online hateful speech. Methods We adopt several Python-based crawlers to collect a comprehensive data set covering a variety of subjects from a multiplicity of online communities in South Korea. We apply the notions of marginalization and polarization in identifying patterns and dynamics of online hateful speech. Results Our analyses suggest that polarization driven by political orientation and age difference predominates in the hateful speech in most communities, while marginalization of social minority groups is also salient in other communities. Furthermore, we identify a temporal shift in the trends of online hate from gender to age based, reflecting the changing sociopolitical conditions within the polarization dynamics in South Korea. Conclusion By expanding our understanding of how hatred shifts and evolves in online communities, our study provides theoretical and practical implications for both researchers and policy-makers.

show abstract

“…In scholarship on the subject, intergroup hatred is most often studied in conjunction with intergroup violence (e.g. ( 51 )); understanding hatred in language, though, requires understanding the myriad ways in which words can be used to accomplish the goals of intergroup hate ( 17 ). We specifically rely on three main operationalizations of outgroup hateful rhetoric: (i) language used to incite genocide ( 52 ) (ii) identity-based prejudicial language ( 53 ), and (iii) dehumanizing language ( 17 ).…”

Section: Overview Of the Present Researchmentioning

confidence: 99%

“…As long as a deeper understanding of why and how individuals and collectives use language for hateful purposes eludes us, a long-term solution to the problem will as well. Based on the idea that the language of hate must be understood in order to be mitigated ( 10 , 17–20 ), the present work aims to situate this usage of language in the broader context in which hate may be motivated, framed, explained, and called for: morality.…”

mentioning

confidence: 99%

The (moral) language of hate

et al. 2023

Self Cite

View full text Add to dashboard Cite

Humans use language toward hateful ends, inciting violence and genocide, intimidating and denigrating others based on their identity. Despite efforts to better address the language of hate in the public sphere, the psychological processes involved in hateful language remain unclear. In this work, we hypothesize that morality and hate are concomitant in language. In a series of studies, we find evidence in support of this hypothesis using language from a diverse array of contexts, including the use of hateful language in propaganda to inspire genocide (Study 1), hateful slurs as they occur in large text corpora across a multitude of languages (Study 2), and hate speech on social-media platforms (Study 3). In post hoc analyses focusing on particular moral concerns, we found that the type of moral content invoked through hate speech varied by context, with Purity language prominent in hateful propaganda and online hate speech and Loyalty language invoked in hateful slurs across languages. Our findings provide a new psychological lens for understanding hateful language and points to further research into the intersection of morality and hate, with practical implications for mitigating hateful rhetoric online.

show abstract

Introducing the Gab Hate Corpus: defining and applying hate-based rhetoric to social media posts at scale

Abstract: Introducing the Gab Hate Corpus: Defining and applying hate-based rhetoric to social media posts at scale. Lang Resources & Evaluation (2022).

Cited by 24 publications

References 63 publications

XMD: An End-to-End Framework for Interactive Explanation-Based Debugging of NLP Models

XMD: An End-to-End Framework for Interactive Explanation-Based Debugging of NLP Models

The hatred of all against all? Evidence from online community platforms in South Korea

The (moral) language of hate

Contact Info

Product

Resources

About