A Instagram (other influencers) dataset of 1,625,526 posts from other influencers, collected by Seungbae Kim and Jinyoung Han (2020). Each record contains post caption text, like count, comment count, video/ad flags, tagged user count, location presence, image dimensions, timestamp, and influencer follower count. The dataset is notable for its subset of 9.7M-post Instagram influencer dataset covering the other category, with rich per-post engagement signals and influencer-level metadata. Intended for research into influencer marketing analysis, engagement prediction, social media content analysis, and sponsorship detection. This package holds the post metadata (one JSON .info file per post) and influencer lists; the post images, which the authors distribute separately on request, are not included.
A Instagram (pet influencers) dataset of 169,998 posts from pet influencers, collected by Seungbae Kim and Jinyoung Han (2020). Each record contains post caption text, like count, comment count, video/ad flags, tagged user count, location presence, image dimensions, timestamp, and influencer follower count. The dataset is notable for its subset of 9.7M-post Instagram influencer dataset covering the pet category, with rich per-post engagement signals and influencer-level metadata. Intended for research into influencer marketing analysis, engagement prediction, social media content analysis, and sponsorship detection. This package holds the post metadata (one JSON .info file per post) and influencer lists; the post images, which the authors distribute separately on request, are not included.
A Instagram (travel influencers) dataset of 1,203,723 posts from travel influencers, collected by Seungbae Kim and Jinyoung Han (2020). Each record contains post caption text, like count, comment count, video/ad flags, tagged user count, location presence, image dimensions, timestamp, and influencer follower count. The dataset is notable for its subset of 9.7M-post Instagram influencer dataset covering the travel category, with rich per-post engagement signals and influencer-level metadata. Intended for research into influencer marketing analysis, engagement prediction, social media content analysis, and sponsorship detection. This package holds the post metadata (one JSON .info file per post) and influencer lists; the post images, which the authors distribute separately on request, are not included.
A Instagram (user engagement patterns) dataset of 1,000 posts with engagement metrics and multimedia, collected by Hanmante via Instagram API (2024). Each record contains post captions, hashtags, images, videos, likes, comments, user metadata, and tagged users. The dataset is notable for its dataset of Instagram political content from Uruguay, with rich multimodal data. Intended for research into social media engagement analysis, political communication, multimodal NLP, and user behavior modeling.
A ISOT (fake news dataset) dataset of 44,898 news articles labeled fake/real, collected by ISOT via Reuters and various sources (2017). Each record contains news article text (title + content) and binary label (is_fake: 1=fake, 0=real). The dataset is notable for its well-balanced dataset of fake and real news articles for misinformation research. Intended for research into fake news detection, text classification, and misinformation analysis.
A Jester Online Joke Recommender System (UC Berkeley) dataset of 73,521 nodes and 4,136,360 edges, collected by K. Goldberg et al. via Jester system. Each record contains directed: False, weighted: True, temporal: False. Bipartite user-joke rating network with continuous ratings from -10 to +10 — one of the earliest large-scale collaborative filtering datasets.. Suitable for recommender systems, collaborative filtering, continuous rating prediction.
A LASAGNE online social network dataset dataset of 12,643 nodes and 57,453 edges, collected by KONECT. Each record contains directed: True, weighted: True. Directed network from the LASAGNE project — a small online social network dataset.. Suitable for social network analysis, community detection.
A Last.fm music streaming and social platform dataset of 174,077 nodes and 894,472 edges, collected by KONECT via Last.fm API. Each record contains directed: False, weighted: True, temporal: True. Temporal bipartite network of Last.fm users and bands they listened to, with Unix timestamps per play event.. Suitable for music recommendation, temporal bipartite networks, listening behavior analysis.
A Last.fm music streaming and social platform dataset of 1,084,620 nodes and 4,413,416 edges, collected by KONECT via Last.fm API. Each record contains directed: False, weighted: True, temporal: True. Temporal bipartite network of Last.fm users and individual songs they listened to, with Unix timestamps per play event.. Suitable for music recommendation, temporal bipartite networks, song preference analysis.
A LIAR (fact-checking benchmark) dataset of 12,789 labeled claim statements from Politifact, collected by Wang via Politifact fact-checking (2017). Each record contains claim text, 6-class veracity label, speaker metadata (party, state, job title), credit history counts (barely_true/false/half_true/mostly_true/pants_fire), and context. The dataset is notable for its benchmark dataset for fake news detection with fine-grained 6-class labels, rich speaker metadata, and credit history. Intended for research into fake news detection, fact-checking, political discourse analysis, and misinformation research.
A Libimseti.cz (Czech online dating platform) dataset of 220,970 nodes and 17,359,346 edges, collected by Brozovsky & Petricek via platform crawl. Each record contains directed: True, weighted: True, signed: False, temporal: False. Directed weighted rating network from a Czech dating site where users rate each other 1-10. One of the largest online dating networks publicly available for research.. Suitable for recommender systems, rating network analysis, online dating research.
A LinkedIn dataset of 3,600 profiles: 1,800 legitimate, 600 fake and 1,200 generated by ChatGPT (600 from legitimate profiles' statistics, 600 from fake ones'), released by Ayoobi, Shahriar and Mukherjee (ACM Hypertext 2023). Each record holds the profile sections visible before connecting: intro, about, experience, education, skills, languages, interests, licences and more, with counts per section. Suitable for fake and LLM-generated profile detection.
A Linux kernel source code (function call graph) dataset of 30,837 nodes and 213,954 edges, collected by KONECT via static analysis of Linux kernel. Each record contains directed: True, weighted: False. Directed call graph of the Linux kernel — nodes are functions, edges are function calls.. Suitable for software network analysis, dependency network research, community detection.
A LiveJournal (blogging social network) dataset of 12,404,868 emotion-labeled posts from 115,444 users with 106,577 nodes and 961,392 friendship/follower edges, collected by Jin & Zafarani via LiveJournal API, distributed by Data Lab SU. Each record contains user ID, emotion label (-1=negative/0=neutral/1=positive), timestamp, friendship edges, follower edges, community membership. 12,404,868 LiveJournal posts labeled with trichotomous emotion scores across 115,444 users — uniquely combines social network structure (friendship + follower graphs) with per-post emotion labels and community membership for studying emotion propagation in networks. Suitable for emotion propagation in social networks, community emotion analysis, sentiment diffusion, social influence on emotion, network-based mood modeling.
A LiveJournal social network of 106,577 users and 961,392 directed edges (793,948 follower links and 246,164 friendship links), from a crawl of the users within three hops of the seed user krazykitkat, pre-processed for the paper 'Emotions in Social Networks: Distributions, Patterns, and Models' (Jin and Zafarani): only users with 10 or more posts carrying a predefined LiveJournal mood were kept, and user ids are anonymized. This copy holds the two network files (Friendship.txt, Followers.txt) and the release README; the per-post mood file (Posts.txt) and community memberships (Community.txt) that the README describes are not included. Suitable for social network structure, follower and friendship analysis, community detection, social influence modeling.
A LiveJournal (blogging and social network platform) dataset of 4,846,609 nodes and 68,475,391 edges, collected by J. Leskovec et al. via LiveJournal API. Each record contains directed: True, weighted: False, temporal: False. Large directed LiveJournal friendship network with 4.3M users and 69M edges — one of the largest social network datasets from the Web 2.0 era.. Suitable for large-scale social network analysis, community detection, information diffusion.
A Livemocha (language learning community) dataset of 104,103 users and 2,193,083 friendship edges, collected by Xia Hu (Ben) via Livemocha crawl, December 2010. Each record contains user-user friendship edges. The dataset is notable for its friendship network of the world's largest online language learning community, with 6M+ members across 200+ countries. Intended for research into social network analysis, community detection, and cross-cultural interaction studies.
A Linux Kernel Mailing List (LKML) — open source development community dataset of 27,927 nodes and 242,976 edges, collected by Ingo Scholtes et al. via mailing list archive. Each record contains directed: True, weighted: True, signed: False, temporal: True. Temporal directed reply network of the Linux Kernel Mailing List — one of the longest-running and most active open source development communities. Edges carry Unix timestamps enabling longitudinal analysis.. Suitable for temporal network analysis, community detection, open source collaboration.
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 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 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 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 FAA air traffic control network dataset of 1,226 nodes and 2,615 edges, collected by A. Ma'ayan et al.. Each record contains directed: True, weighted: False. FAA air traffic control network from the Ma'ayan lab dataset collection.. Suitable for infrastructure network analysis, biological network benchmarking.
A Human protein interaction network (Figeys lab) dataset of 2,239 nodes and 6,452 edges, collected by D. Figeys et al.. Each record contains directed: True, weighted: False. Human protein-protein interaction network from the Figeys lab.. Suitable for biological network analysis, protein interaction research.
A Ecological food web (Ma'ayan lab collection) dataset of 183 nodes and 2,494 edges, collected by A. Ma'ayan et al.. Each record contains directed: True, weighted: False. Ecological food web network from the Ma'ayan lab dataset collection.. Suitable for ecological network analysis, food web topology.
A PDZBase protein domain interaction database dataset of 212 nodes and 244 edges, collected by A. Ma'ayan et al.. Each record contains directed: False, weighted: False. PDZ domain protein-protein interaction network from the PDZBase database.. Suitable for biological network analysis, domain interaction research.
A Human protein interaction network (Stelzl et al.) dataset of 1,706 nodes and 6,207 edges, collected by U. Stelzl et al.. Each record contains directed: True, weighted: False. Human protein-protein interaction network from Stelzl et al. — yeast two-hybrid screening.. Suitable for biological network analysis, protein interaction research.
A Human protein interaction network (Vidal lab) dataset of 3,133 nodes and 6,726 edges, collected by M. Vidal et al.. Each record contains directed: False, weighted: False. Human protein-protein interaction network from the Vidal lab — classic biological network benchmark.. Suitable for biological network analysis, protein interaction research.
A MADOC (Bluesky, bluesky) dataset of 164,265 users and 894,917 reply edges from 2,814,352 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Koo, koo) dataset of 124,477 users and 1,121,958 reply edges from 4,263,640 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, CringeAnarchy) dataset of 202,705 users and 1,864,741 reply edges from 6,445,396 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, KotakuInAction) dataset of 83,581 users and 1,758,653 reply edges from 6,793,795 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, MensRights) dataset of 113,904 users and 822,329 reply edges from 3,110,425 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, fatpeoplehate) dataset of 1,409,008 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, funny) dataset of 3,851,024 users and 21,255,711 reply edges from 61,859,095 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Suitable for social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, gaming) dataset of 2,870,130 users and 14,950,380 reply edges from 44,274,130 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, gifs) dataset of 1,894,135 users and 6,785,726 reply edges from 21,090,022 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, greatawakening) dataset of 18,267 users and 302,678 reply edges from 915,265 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, milliondollarextreme) dataset of 19,187 users and 422,964 reply edges from 1,286,148 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, pics) dataset of 3,453,264 users and 15,863,415 reply edges from 51,914,975 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Suitable for social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, technology) dataset of 611,131 users and 2,052,369 reply edges from 11,766,082 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Reddit, videos) dataset of 1,854,456 users and 8,653,524 reply edges from 36,775,460 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, CringeAnarchy) dataset of 876 users and 1,740 reply edges from 3,030 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, KotakuInAction) dataset of 1,878 users and 3,854 reply edges from 7,853 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, MensRights) dataset of 992 users and 1,628 reply edges from 3,735 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, fatpeoplehate) dataset of 20,942 users and 226,749 reply edges from 352,003 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, funny) dataset of 17,128 users and 82,319 reply edges from 143,945 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, gaming) dataset of 9,794 users and 28,073 reply edges from 69,478 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
A MADOC (Voat, gifs) dataset of 5,022 users and 11,435 reply edges from 21,573 interactions, collected by Dankulov et al. via platform APIs, published at ICWSM 2025. Each record contains user-user reply edges, post and comment text, per-record sentiment scores (VADER and TextBlob), subjectivity scores, and toxicity scores (ToxiGen). The dataset is notable for its FAIR-compliant dataset with pre-computed NLP quality labels (sentiment, subjectivity, toxicity) per record, enabling network analysis and content quality research without additional annotation. Intended for research into social network analysis, toxicity detection, cross-platform community dynamics, misinformation research, and hate speech propagation studies.
Instagram
Last.fm
Libimseti
LinkedIn
GitHub
LiveJournal
Livemocha
4chan
Twitter (X)
Weibo
Bluesky
Koo
Reddit
Voat