A Voat (Reddit alternative, 2013–2020) dataset of 165,864 submissions, 1,485,665 comments, 73,700 users, 727,833 collaboration edges, collected by Mekacher & Papasavva (ICWSM 2022) via full platform crawl. Each record contains user nodes (registration date, moderation status), directed weighted edges (commenter→submitter, weight=upvotes+1), submission titles and bodies, comment text, upvotes, downvotes, depth. Aggregate of Voat communities containing explicit hate speech targeting racial, religious, and other minority groups. Included for research transparency — enables hate speech detection, toxicity modeling, and online radicalization studies using real platform data.. Suitable for hate speech detection, toxicity modeling, online radicalization, content moderation research, platform dynamics.
A 4chan (/pol/) dataset of 499,723 posts with moderation scores, collected by kjj0 via 4chan API and OpenAI moderation model. Each record contains post text, sexual score, hate score, violence score, self-harm score, sexual/minors score, hate/threatening score, violence/graphic score. A dataset of 500,000 /pol/ posts with OpenAI moderation scores, enabling research on content moderation, hate speech detection, and harmful content analysis. Suitable for content moderation, hate speech detection, toxicity analysis, misinformation, political communication.
A Twitter (hateful user network) dataset of 100,386 users (544 hateful, 4,427 normal, 0 suspended) with 2,286,592 follower edges, collected by Ribeiro et al. via Twitter API (2018). Each record contains user ID, hate label (hateful/normal/suspended), follower/followee/statuses counts, network centrality metrics, 200+ Empath emotional category scores, 300-dim GloVe embeddings, sentiment, subjectivity, tweet engagement statistics. One of the most feature-rich hateful user datasets, with 100,386 Twitter users labeled for hatefulness and supplemented with full network topology, Empath psychological features, and GloVe embeddings. Suitable for hate speech detection, hateful user classification, network-based toxicity analysis, suspended account detection, social influence of hate, graph-based hate propagation.
A Multi-Platform (Social Media) dataset of 26,985 bilingual (Malay-English) social media posts, collected by Mendeley Data. Each record contains text, label (hate/non-hate), language. Balanced bilingual Malay-English hate speech dataset with cleaned text. Suitable for hate speech detection, multilingual NLP, content moderation.
A Twitter (crowdsourced hate speech annotation) dataset of 24,783 tweets labeled as hate speech (1,430), offensive (19,190), or neither (4,163), collected by Davidson et al. via CrowdFlower crowdsourcing (2017). Each record contains tweet text, majority class label (0=hate/1=offensive/2=neither), per-label coder counts, total coder count. Davidson et al. 2017 is one of the most cited hate speech datasets, with 24,783 tweets annotated by multiple crowdworkers — distinguished by separating hate speech from merely offensive language. Suitable for hate speech detection, offensive language classification, content moderation, NLP benchmarking, crowdsourced annotation research.
A Twitter (#BlackLivesMatter) dataset of 9,165 tweets (3,084 hate speech, 6,081 non-hate), collected by Giorgi et al. via Twitter API (2020). Each record contains tweet text, binary hate label (0/1), hashtags, source platform, user friends/favourites count, retweet flag. 9,165 BLM-related tweets with hate speech labels — captures online discourse during the 2020 Black Lives Matter movement with 3,084 hate speech examples. Suitable for hate speech detection, social movement analysis, BLM discourse research, Twitter content moderation, political communication, racial justice NLP.
A HOT Speech (hate/offensive/toxic speech dataset) dataset of 3,481 social media comments with multi-dimensional annotations, collected by SOMAR / ICPSR via Twitter, YouTube, and Reddit APIs (August 2021). Each record contains comment text, hard and soft labels (hateful, offensive, toxic), composite scores, and 21 researcher codes. multi-platform dataset with rich annotations including both MTurk worker labels and researcher codes. Suitable for hate speech detection, offensive language identification, toxicity classification, and political discourse analysis.
A 4chan, Twitter, and Weibo (cross-platform) dataset of 2,455 multimodal posts (1,318 hate, 1,137 non-hate), collected by Mira AI Lab (2024). Each record contains image filename, image OCR text, post text, hate/non-hate binary label, hate category (gender/race/religion/politics/violence/health/etc), and human-annotated reason for classification. Multimodal cross-platform collection hate speech dataset combining image OCR text and post text with fine-grained category labels across 8 hate dimensions — 54% of posts are hate speech with human-annotated reasoning. Suitable for multimodal hate speech detection, cross-platform analysis, hate category classification, image-text hate speech, content moderation research. The package includes the 2,455 meme images (img/) that the records name.
A 4chan dataset of 1,400 multimodal posts (1,062 hate, 338 non-hate), collected by Mira AI Lab (2024). Each record contains image filename, image OCR text, post text, hate/non-hate binary label, hate category (gender/race/religion/politics/violence/health/etc), and human-annotated reason for classification. Multimodal 4chan subset hate speech dataset combining image OCR text and post text with fine-grained category labels across 8 hate dimensions — 76% of posts are hate speech with human-annotated reasoning. Suitable for multimodal hate speech detection, cross-platform analysis, hate category classification, image-text hate speech, content moderation research. The package includes the 1,400 meme images (img/) that the records name.
A NoisyHate (hate speech benchmark) dataset of 1,339 labeled sentences with text perturbations and toxicity scores, collected by NoisyHate Authors via ICWSM 2025. Each record contains sentence text, binary hate label, toxicity scores (toxicity, severe, obscene, identity_attack, insult, threat), identity group indicators, and engagement metrics. The dataset is notable for its benchmark dataset with human-written text perturbations and fine-grained toxicity metrics. Intended for research into hate speech detection, model robustness, adversarial testing, toxicity analysis, and identity-based hate speech research.
A Cross-platform meme corpus (4chan, Twitter, Weibo) dataset of 1,318 hate-speech labeled meme records from 3 platforms, collected by Mira AI Lab via cross-platform meme collection. Each record contains image filename, OCR-extracted image text, post caption text, hate speech label, category list (politics/race/religion/etc.), and human-readable annotation reason per record. M3-Hate is the hate-speech filtered subset of the M3 meme dataset, with 1,318 hateful memes across 4chan (1062), Twitter (132), and Weibo (124) — multimodal with both image OCR text and post captions. Suitable for hate speech detection, multimodal meme analysis, cross-platform content moderation, image-text classification, harmful content research. The package includes the 1,318 meme images (img/) alongside the JSON records.
A Weibo dataset of 529 multimodal posts (124 hate, 405 non-hate), collected by Mira AI Lab (2024). Each record contains image filename, image OCR text, post text, hate/non-hate binary label, hate category (gender/race/religion/politics/violence/health/etc), and human-annotated reason for classification. Multimodal Weibo subset hate speech dataset combining image OCR text and post text with fine-grained category labels across 8 hate dimensions — 23% of posts are hate speech with human-annotated reasoning. Suitable for multimodal hate speech detection, cross-platform analysis, hate category classification, image-text hate speech, content moderation research. The package includes the 529 meme images (img/) that the records name.
A Twitter/X dataset of 526 multimodal posts (132 hate, 394 non-hate), collected by Mira AI Lab (2024). Each record contains image filename, image OCR text, post text, hate/non-hate binary label, hate category (gender/race/religion/politics/violence/health/etc), and human-annotated reason for classification. Multimodal Twitter subset hate speech dataset combining image OCR text and post text with fine-grained category labels across 8 hate dimensions — 25% of posts are hate speech with human-annotated reasoning. Suitable for multimodal hate speech detection, cross-platform analysis, hate category classification, image-text hate speech, content moderation research. The package includes the 526 meme images (img/) that the records name.
Voat
4chan
Twitter (X)
Instagram
Reddit
YouTube
Weibo