Dataset of Arabic religious hate tweets sampled using neutral religious names as keywords. Annotation was crowdsourced using CrowdFlower, with a minimum of 3 annotations per...

Dataset of sexist tweets sampling based on benevolent sexist key phrases from which 712 tweets were manually selected by the authors, and validated by three non-activist...

Dataset of racist and sexist tweets sampled from Twitter and labelled first by experts (including feminist and anti-racist activists), and then by CF amateur annotators who...

Dataset of 1528 annotated comments from Fox News website, taken from 10 news articles. Comments were labelled by experts in two stages, with an annotator agreement of Cohen's...

Dataset of German annotated corpus of tweets regarding refugees in Germany. Tweets were sampled using 10 hateful hashtags and labelled by experts with 2 annotators per tweet....

Dataset of hateful tweets against immigrants, roma and muslims, sampled using keywords. 3,154 tweets were annotated by experts (2 per tweet) and then 2,855 annotated by CF (3+...

Dataset of offensive comments by collecting comments in the Brazilian Web from g1.gobo.com, the most accessed news site in Brazil. Sampled by comments about politics and sports...

Dataset of pairs islamophobic hate speech and counter-responses with 3 types of metadata: expert demographics, hate speech sub-topic, counter-narrative type. The dataset is...

Dataset of hateful/ not hateful posts in the context of conversations from Reddit and Gab. The data is annotated through crowd-sourcing with Amazon Mechanical Turk with...

Dataset of racist and sexist tweets sampled from Twitter and labelled by a mix of expert annotators and activists. Tweets were sampled in 2016 over 2 months using keywords....

Hate Speech Dataset for Hindi, German and English. Three datasets sampled from Twitter and Facebook sampled by topics, hashtags, other keywords and the timeline of users (last...

Hate speech dataset composed of thousands of sentences extracted from Stormfront, a white supremacist forum, manually labelled by experts. Annotator agreement for the 1st round...

Dataset of multi-aspect hate speech posts sampled from Twitter and labelled through crowd-sourcing (Amazon Mechanical Turk). Tweets were sampled by common slurs and demeaning...

Dataset of hate speech and targets from Twitter collected through a multi-step classification process and annotated through CrowdFlower. 92.8% agreement among the annotators for...

10 annotators initially annotated 20,000 entries from the Everyday Sexism Project. At least two annotators labelled every entry. Average Cohen's kappa for the per-category pairs...

Task description: Branching structure of tasks: Binary (Offensive, Not), Within Offensive (Target, Not), Within Target (Individual, Group, Other) Details of task:...

The Turkish dataset used in OffensEval 2020