Twitter History and Image Sharing (THIS) v1.0: country-level aggregates of about 14.6 billion geolocated tweets, 2013 to 2023, released by Zachary Steinert-Threlkeld on Harvard Dataverse. Four tables, 17,928,584 rows in all: tweets and users per country and day (792,313 rows), tweets per country, language and day (15,407,849), new accounts per country and day (943,361), and image-sharing availability per country and day (785,061). No individual tweets or users are included. Suitable for country-level trends in platform use, language use and image sharing.
A X (Twitter) with Community Notes annotations dataset of 152,266 multimodal posts (21,845 AI-generated, 104,915 miscaptioned), collected by Chrysidis, Papadopoulos & Papadopoulos via X Community Notes (2026). Each record contains tweet ID, reply/favorite/view/retweet counts, Community Notes text and timestamp, and misinformation type label (miscaptioned / ai_generated / other). Large-scale dataset of 152,266 multimodal X posts with Community Notes — tracking the rise of AI-generated misinformation with 21,845 confirmed AI-generated posts. Includes engagement metrics showing that AI-generated content achieves disproportionate virality despite slower initial reporting.. Suitable for AI-generated content detection, misinformation classification, Community Notes effectiveness, multimodal misinformation analysis, virality and engagement modeling, temporal degradation of detectors.
A Twitter dataset of 16,057 tweets across 7 crisis events referencing 17,465 images by URL, collected by Alam et al. for 7 major natural disasters in 2017. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its Systematically collected multimodal dataset of 7 natural disasters with full Twitter API metadata. Cumulative version combines all events into a single corpus. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, cross-event engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
A Twitter dataset of 4,021 tweets referencing 3,611 images by URL, collected by Alam et al. for the Hurricane Irma 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the Hurricane Irma 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
A Twitter dataset of 3,995 tweets referencing 4,691 images by URL, collected by Alam et al. for the Hurricane Maria 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the Hurricane Maria 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
A Twitter dataset of 3,992 tweets referencing 4,569 images by URL, collected by Alam et al. for the Hurricane Harvey 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the Hurricane Harvey 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
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 Twitter dataset of 1,486 tweets referencing 1,590 images by URL, collected by Alam et al. for the California Wildfires 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the California Wildfires 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
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 Twitter dataset of 1,237 tweets referencing 1,380 images by URL, collected by Alam et al. for the Mexico Earthquake 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the Mexico Earthquake 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
A Twitter dataset of 830 tweets referencing 1,027 images by URL, collected by Alam et al. for the Southeast Asia Floods 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the Southeast Asia Floods 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
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.
A Twitter dataset of 496 tweets referencing 597 images by URL, collected by Alam et al. for the Iraq-Iran Earthquake 2017 crisis. Each record contains tweet text, engagement metrics (retweets, favorites, replies, quotes), user metadata, media attachments, crisis code. The dataset is notable for its multimodal dataset from the Iraq-Iran Earthquake 2017 crisis, collected from Twitter during the event with full Twitter API metadata. Intended for research into crisis informatics, multimodal learning, disaster response, image classification, tweet filtering, engagement analysis. This copy (the SocialED redistribution) holds the tweet objects only: the images are referenced by URL and the image files are not included, and neither are CrisisMMD's human annotations (informativeness, humanitarian category, damage severity).
4chan
Weibo