Automated Hate Speech Detection and the Problem of Offensive Language

Bibliographic Data

ID23328286
AuthorsThomas R Davidson (0000-0002-5947-7490, Cornell University), Thomas Davidson (0000-0002-9393-2573), Dana Warmsley (Cornell University), Michael W Macy (0000-0003-0024-5027, Cornell University), Michael Macy, Ingmar Weber (0000-0003-4169-2579, Hamad bin Khalifa University)
Year2017
Volume11
Issue1
Pages512-515
Publication date2017-05-03
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueProceedings of the International AAAI Conference on Web and Social Media (JOURNAL)
Journal identifiersISSN: 2162-3449 • E-ISSN: 2334-0770
PublisherAssociation for the Advancement of Artificial Intelligence (AAAI) (PUBLISHER)
DOI10.1609/icwsm.v11i1.14955
OpenAlexW2595653137
LanguageEN
Citations received140
References cited4

A key challenge for automatic hate-speech detection on social media is the separation of hate speech from other instances of offensive language. Lexical detection methods tend to have low precision because they classify all messages containing particular terms as hate speech and previous work using supervised learning has failed to distinguish between the two categories. We used a crowd-sourced hate speech lexicon to collect tweets containing hate speech keywords. We use crowd-sourcing to label a sample of these tweets into three categories: those containing hate speech, only offensive language, and those with neither. We train a multi-class classifier to distinguish between these different categories. Close analysis of the predictions and the errors shows when we can reliably separate hate speech from other offensive language and when this differentiation is more difficult. We find that racist and homophobic tweets are more likely to be classified as hate speech but that sexist tweets are generally classified as offensive. Tweets without explicit hate keywords are also more difficult to classify.

Classifier (UML) · Lexicon · Natural language processing · Offensive · Speech processing · Speech recognition · Voice activity detection · Artificial Intelligence · Computer Science · Hate Speech and Cyberbullying Detection · Internet Traffic Analysis and Secure E-voting · Mathematics · Spam and Phishing Detection

Tools

Open DOIOpen Access