A Weibo (CHECKED dataset) dataset of 2,104 microblogs with engagement and expert analysis, collected by Yang, Zhou, and Zafarani via Weibo (2020-2021). Each record contains microblog text, images, videos, engagement metrics (comments, reposts, likes), and expert analysis for fake news. The dataset is notable for its first Chinese COVID-19 fake news dataset with expert analysis, anonymized user IDs, and rich social interaction data. Intended for research into fake news detection, misinformation analysis, Chinese NLP, social network analysis, and public health communication.
A TextDetox 2024 shared task (Hebrew) dataset of 2,011 Hebrew posts (807 toxic, 1,204 non-toxic), collected by Dementieva et al. for PAN/CLEF TextDetox 2024. Each record contains post text, binary toxicity label (0=non-toxic, 1=toxic). TextDetox Hebrew provides 2,011 balanced binary toxicity annotations — part of a 15-language multilingual benchmark for cross-lingual toxicity detection research. Suitable for toxicity detection, multilingual content moderation, cross-lingual NLP, hate speech classification, text detoxification.
A Blogger.com (2004) dataset of 1,917 bloggers, 2,293 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
A UC Irvine online messaging system dataset of 1,899 nodes and 20,296 edges, collected by T. Opsahl. Each record contains directed: True, weighted: True, signed: False. Temporal directed messaging network among UC Irvine students with Unix timestamps.. Suitable for temporal network analysis, online social network analysis. The file holds 59,835 messages, one row per message; the edge count of 20,296 counts distinct pairs.
A Protein-protein interaction network dataset of 1,870 nodes and 2,277 edges, collected by KONECT. Each record contains directed edges: False, weighted: False, signed: False. This undirected network contains protein interactions contained in yeast. Research showed that proteins with a high degree were more important for the surivial of the yeast than others. A node represents a protein and an edge represents a metabolic interaction between two proteins. The network cont. Suitable for biological network analysis, community detection.
A Petster.com (pet social network — hamsters) dataset of 1,858 nodes and 12,534 edges, collected by KONECT via Petster crawl. Each record contains directed: False, weighted: False, signed: False. Friendship network of hamster profiles on a pet social networking site.. Suitable for social network analysis, community detection.
A Blogger.com (2004) dataset of 1,804 bloggers, 2,070 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
A Name association network dataset of 1,773 nodes and 9,131 edges, collected by J.L. Moreno. Each record contains directed edges: False, weighted: True, signed: False. This undirected network contains nouns (places and names) of the King James Version of the Bible and information about their co-occurrences. A node represents one of the above noun types and an edge indicates that two nouns appeared together in the same Bible verse. The edge multiplicity denotes ho. Suitable for semantic network analysis.
A CLEF CheckThat! 2025 Task 2 (portuguese) dataset of 1,735 social media post and normalized claim pairs, collected by CLEF CheckThat! Lab organizers (2025). Each record contains social media post text containing a factual claim, normalized canonical claim statement. CLEF 2025 CheckThat! Task 2 (portuguese) provides 1,735 pairs of social media posts and their normalized factual claims, enabling claim normalization and canonicalization research. Suitable for claim normalization, fact-checking automation, NLP text generation, multilingual claim extraction, social media fact verification.
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 Blogger.com (2004) dataset of 1,588 bloggers, 1,861 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
A Chicago regional road network (TNTP benchmark) dataset of 1,467 nodes and 1,298 edges, collected by Transportation Networks for Research (TNTP). Each record contains directed: False, weighted: False. Road network of the Chicago metropolitan area — classic transportation network benchmark.. Suitable for transportation network analysis, routing algorithms, network flow.
A UC Irvine online student forum dataset of 1,421 nodes and 7,089 edges, collected by T. Opsahl. Each record contains directed: False, weighted: True, signed: False. Temporal bipartite network of students posting to forum threads with Unix timestamps.. Suitable for temporal bipartite network analysis, online community analysis.
A Blogger.com (2004) dataset of 1,376 bloggers, 1,657 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
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 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 Political blogosphere (US 2004 election) dataset of 1,224 nodes and 19,025 edges, collected by Adamic & Glance via blog crawl. Each record contains directed edges: True, weighted: False, signed: False. This directed network contains front-page hyperlinks between blogs in the context of the 2004 US election. A node represents a blog and an edge represents a hyperlink between two blogs.. Suitable for political polarization, community detection, information diffusion.
A CLEF CheckThat! 2025 Task 2 (french) dataset of 1,174 social media post and normalized claim pairs, collected by CLEF CheckThat! Lab organizers (2025). Each record contains social media post text containing a factual claim, normalized canonical claim statement. CLEF 2025 CheckThat! Task 2 (french) provides 1,174 pairs of social media posts and their normalized factual claims, enabling claim normalization and canonicalization research. Suitable for claim normalization, fact-checking automation, NLP text generation, multilingual claim extraction, social media fact verification.
The international E-road network, a road network located mostly in Europe, compiled by Lovro Šubelj and Marko Bajec and distributed by KONECT (subelj_euroroad): 1,174 cities linked by 1,417 undirected roads, where an edge means two cities are joined by an E-road. City names are included. Useful for studying spatial and infrastructure networks, network robustness and community detection.
A University Rovira i Virgili (email communication network) dataset of 1,133 nodes and 5,451 edges, collected by R. Guimerà et al.. Each record contains directed: False, weighted: False, signed: False. Email communication network of a university — classic benchmark for community detection algorithms.. Suitable for community detection, email network analysis, graph algorithm benchmarking.
A CLEF CheckThat! 2025 Task 2 (hindi) dataset of 1,081 social media post and normalized claim pairs, collected by CLEF CheckThat! Lab organizers (2025). Each record contains social media post text containing a factual claim, normalized canonical claim statement. CLEF 2025 CheckThat! Task 2 (hindi) provides 1,081 pairs of social media posts and their normalized factual claims, enabling claim normalization and canonicalization research. Suitable for claim normalization, fact-checking automation, NLP text generation, multilingual claim extraction, social media fact verification.
A CLEF CheckThat! 2024 shared task (dutch) dataset of 995 sentences with binary check-worthiness labels (40.7% check-worthy), collected by CLEF CheckThat! Lab organizers (2024). Each record contains sentence text, binary check-worthiness label (Yes/No), sentence ID, source document. CLEF 2024 CheckThat! Task 1 (dutch) provides 995 sentences from political debates and social media annotated for check-worthiness — whether the claim merits fact-checking. Suitable for check-worthiness detection, fact-checking automation, misinformation research, NLP classification benchmarks, political discourse analysis.
A European research institution internal email system dataset of 986 employees and 332,334 directed email interactions across 42 departments, collected by Leskovec et al. via SNAP (anonymized institutional email logs). Each record contains sender employee ID, recipient employee ID, Unix timestamp, and department label per employee (42 departments). Temporal directed email communication network from a European research institution with ground-truth department labels for all 986 employees — enabling evaluation of community detection algorithms against known organizational structure.. Suitable for community detection benchmarking, organizational network analysis, temporal email network analysis, higher-order graph clustering.
A Democratic National Committee (email communications, 2016) dataset of 906 nodes and 10,429 edges, collected by WikiLeaks DNC email leak. Each record contains directed: False, weighted: True, signed: False, temporal: False. Weighted undirected co-recipient network from the leaked DNC emails — two people are connected if they appeared together on email threads, with edge weights indicating co-occurrence frequency.. Suitable for political network analysis, email network analysis, community detection.
A Terrorist Relations dataset of 881 nodes, 8,592 edges, 4 classes, collected by Yang et al. (ICML 2016). Each record contains node features, citation edges, class labels. Terrorist Relations network with bag-of-words features. Suitable for node classification, semi-supervised learning, graph neural networks.
A Unicode CLDR language data dataset of 868 nodes and 1,255 edges, collected by KONECT via Unicode CLDR. Each record contains directed: False, weighted: True. Bipartite weighted network of languages and Unicode scripts, with edge weights representing similarity scores.. Suitable for linguistic network analysis, language typology, script similarity.
A Terrorist Attacks dataset of 645 nodes, 3,172 edges, 6 classes, collected by Yang et al. (ICML 2016). Each record contains node features, citation edges, class labels. Terrorist Attacks network with bag-of-words features. Suitable for node classification, semi-supervised learning, graph neural networks.
A Blogger.com (2004) dataset of 570 bloggers, 906 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
A Blogger.com (2004) dataset of 557 bloggers, 686 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
A Blogger.com (2004) dataset of 541 bloggers, 678 posts, collected by Schler et al. for AAAI 2006. Each record contains blog post text, gender, age, industry, astrological sign. Historical blog corpus from 2004 with rich metadata per blogger (gender, age, industry, astrological sign). Balanced across age groups and genders.. Suitable for authorship attribution, stylistic analysis, demographic language variation, historical NLP.
A CLEF CheckThat! 2025 Task 2 (arabic-msa) dataset of 540 social media post and normalized claim pairs, collected by CLEF CheckThat! Lab organizers (2025). Each record contains social media post text containing a factual claim, normalized canonical claim statement. CLEF 2025 CheckThat! Task 2 (arabic-msa) provides 540 pairs of social media posts and their normalized factual claims, enabling claim normalization and canonicalization research. Suitable for claim normalization, fact-checking automation, NLP text generation, multilingual claim extraction, social media fact verification.
A CLEF CheckThat! 2025 Task 2 (arabic) dataset of 470 social media post and normalized claim pairs, collected by CLEF CheckThat! Lab organizers (2025). Each record contains social media post text containing a factual claim, normalized canonical claim statement. CLEF 2025 CheckThat! Task 2 (arabic) provides 470 pairs of social media posts and their normalized factual claims, enabling claim normalization and canonicalization research. Suitable for claim normalization, fact-checking automation, NLP text generation, multilingual claim extraction, social media fact verification.
A C. elegans metabolic network dataset of 453 nodes and 2,040 edges, collected by H. Jeong et al.. Each record contains directed: False, weighted: True, signed: False. Metabolic network of C. elegans — classic biological network benchmark.. Suitable for biological network analysis, community detection, scale-free network research.
A CLEF CheckThat! 2025 Task 2 (punjabi) dataset of 445 social media post and normalized claim pairs, collected by CLEF CheckThat! Lab organizers (2025). Each record contains social media post text containing a factual claim, normalized canonical claim statement. CLEF 2025 CheckThat! Task 2 (punjabi) provides 445 pairs of social media posts and their normalized factual claims, enabling claim normalization and canonicalization research. Suitable for claim normalization, fact-checking automation, NLP text generation, multilingual claim extraction, social media fact verification.
A Science Gallery Dublin (Infectious exhibition, 2009) dataset of 410 nodes and 2,765 edges, collected by SocioPatterns collaboration via RFID proximity sensing. Each record contains directed: False, weighted: True, signed: False, temporal: True. High-resolution face-to-face contact network of 410 visitors at a science exhibition, recorded via RFID sensors at 20-second resolution. One of the canonical SocioPatterns datasets for epidemic modeling.. Suitable for epidemic modeling, contact network analysis, temporal networks. The file holds 17,298 contacts, one row per 20-second face-to-face contact; the edge count of 2,765 counts distinct pairs.
Weibo
Blogger
Petster