A Feature-Based Approach to Assess Hate Speech in User Comments

Dados Bibliográficos

ID21761859
AutoresLiane Reiners (Johannes Gutenberg University Mainz), Christian Schemer (0000-0002-7808-2240, Johannes Gutenberg University Mainz)
Ano2020
Volume38
Páginas529-548
Data de publicação2020-01-01
Peer ReviewedSim
Open AccessSim
TipoARTICLE
PeriódicoQuestions de communication (JOURNAL)
Identificadores do periódicoISSN: 1633-5961 • E-ISSN: 2259-8901
EditoraOpenEdition (PUBLISHER)
DOI10.4000/questionsdecommunication.24808
OpenAlexW3186885973
IdiomaEN
Citações recebidas2
Referências citadas35

Hate speech in user comments is a serious challenge for media actors, platforms, legal authorities, and the general public alike. This form of communication can impair the quality of online discussions, but also poison public discourse and civic life. This explains the plethora of studies from different disciplines addressing this topic. However, the diversity of disciplines and approaches complicate a common understanding of hate speech. Frequently, definitions are too broad, as they include factors such as the intention of the commenter or the consequences for stigmatized groups. Additionally, binary categorizations (hate/no hate) applied in some disciplines fail to consider that hate speech varies gradually and ignore qualitative differences of hate speech dimensions. The present research suggests a feature-based approach to analyze hate speech in user comments focusing on manifest features (group-related labels, swear words, trait and action attributions as well as treatment recommendations), and demonstrates its use in a pilot study

Attribution · Authorship attribution · Linguistics · Political science · Trait · Bullying, Victimization, and Aggression · Computer Science · Hate Speech and Cyberbullying Detection · Law · Psychology · Social Psychology · Swearing, Euphemism, Multilingualism · Artificial Intelligence

Ferramentas

Abrir DOIOpen Access