Skip to content
BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
LICENSE
LICENSE
 
 
README.md
README.md
 
 

Repository files navigation

🌟 Awesome Graph Datasets 🌟

Awesome License: MIT

📌 We are actively collecting openly available graph datasets of diverse types for research purposes. This repository will be updated regularly.

Contents

📑 Dataset List

Plain Graphs

Name #nodes #edges #labels Type URL
PPI 3,890 76,584 50 undirected [raw] [raw] [preprocessed]
Blogcatalog3 10,312 333,983 39 undirected [raw] [raw] [preprocessed]
Flickr 80,513 5,899,882 195 undirected [raw] [raw] [preprocessed]
Amazon 334863 925872 100 undirected [raw] [preprocessed]
DBLP 425957 1049866 100 undirected [raw] [preprocessed]
Youtube 1,138,499 2,990,443 47 undirected [raw] [preprocessed]
TWeibo 2,320,895 50,655,143 100 directed [raw] [preprocessed]
Orkut 3,072,441 117,185,084 100 undirected [raw] [preprocessed]
LiveJournal 3997962 34681189 100 undirected [raw] [preprocessed]
In-2004 1,382,908 16,539,643 - directed [raw] [preprocessed]
DBLP 5,425,963 17,298,032 - undirected [raw] [preprocessed]
Pokec 1,632,803 30,622,564 - directed [raw] [preprocessed]
LiveJournal 4,847,571 68,475,391 - directed [raw] [preprocessed]
IT-2004 41,291,594 1,135,718,909 - directed [raw] [preprocessed]
Twitter 41,652,230 1,468,365,182 - directed [raw] [preprocessed]
Friendster-small 7,944,949 447,219,610 100 undirected [raw] [raw] [preprocessed]
Friendster 65,608,366 1,806,067,135 100 undirected [raw] [raw] [preprocessed]
OAG 67,768,244 895,368,962 19 undirected [raw] [preprocessed]
UK-2007 105,896,555 3,738,733,648 - directed [raw][preprocessed]
UK-union 133,633,040 5,475,109,924 - directed [raw] [preprocessed]
ClueWeb12 978,408,098 42,574,107,469 - directed [raw]
ClueWeb09 1,684,868,322 7,939,635,651 - directed [raw] [preprocessed]

Welcome to cite our paper if you publish results based on our preprocessed datasets.

@article{yang13homogeneous,
  title={Homogeneous Network Embedding for Massive Graphs via Reweighted Personalized PageRank},
  author={Yang, Renchi and Shi, Jieming and Xiao, Xiaokui and Yang, Yin and Bhowmick, Sourav S},
  journal={Proceedings of the VLDB Endowment},
  volume={13},
  number={5},
  pages={670--683},
  year={2020},
  publisher={VLDB Endowment}
}

@article{shi13realtime,
  title={Realtime Index-Free Single Source SimRank Processing on Web-Scale Graphs},
  author={Shi, Jieming and Jin, Tianyuan and Yang, Renchi and Xiao, Xiaokui and Yang, Yin},
  journal={Proceedings of the VLDB Endowment},
  volume={13},
  number={7},
  pages={966--980},
  year={2020},
  publisher={VLDB Endowment}
}

Temporal Graphs

Signed Graphs

Name #nodes #positive-edges #negative-edges #attributes #labels Bipartite Notes Source
Epinions 131,828 717,667 123,705 - - No SNAP raw directed edges; also matches the sign2vec paper table. SNAP UOI paper
Wikipedia 7,118 83,962 23,118 - - No SNAP wikiElec signed votes only; 6,960 neutral votes excluded from signed edges. SNAP UOI
Slashdot 82,140 425,072 124,130 - - No SNAP raw directed edges. SNAP UOI
Bitcoin OTC 5,881 32,029 3,563 - - No SNAP weighted ratings binarized by rating sign. SNAP
WikiSigned 7,220 83,717 28,422 - - No sign2vec Wiki/KONECT-style preprocessed count. KONECT paper
Reddit 13,749 26,413 34,106 29 edge features - No UOI RedditGraphs all-new graph; SNAP Reddit Hyperlink is a different 55,863-node / 858,490-edge temporal attributed subreddit graph. SNAP Kaggle UW UOI
ADJNet 4,579 10,708 7,044 - 2 No Undirected adjective network; binary node labels reported in the paper. paper
SCOTUS 28,305 43,781 42,102 - 2 No Directed Supreme Court citation network; binary node labels reported in the paper. paper
Bonanza 9,892 35,805 738 - - Yes 7,919 users + 1,973 sellers; raw signed-bipartite file count. data SBGNN paper
U.S. Senate 1,201 14,979 12,104 - - Yes 145 legislators + 1,056 bills; raw signed-bipartite file count. data SBGNN paper
U.S. House 1,796 61,720 52,658 - - Yes 515 legislators + 1,281 bills; raw signed-bipartite file count. data SBGNN paper
Review 486 464 706 - - Yes Final Review repository count: 182 reviewers + 304 papers. SBGCL SBGNN paper
MovieLens-1M 9,992 575,281 424,928 - - Yes 6,040 users + 3,952 movies; SBGCL split data count. SBGCL paper
Amazon-Book 73,857 1,579,974 380,700 - - Yes 35,736 users + 38,121 items; SBGCL split data count. SBGCL paper
Amazon-CDs 97,731 731,734 163,532 - - Yes 51,267 users + 46,464 items; SIGformer ratings >=4 treated as positive. SIGformer
Amazon-Music 5,970 40,043 9,832 - - Yes 3,472 users + 2,498 items; SIGformer ratings >=4 treated as positive. SIGformer
KuaiRec 4,738 36,560 217,423 - - Yes 1,411 users + 3,327 items; SIGformer binary interaction labels. SIGformer
KuaiRand 21,347 116,679 146,421 - - Yes 16,974 users + 4,373 items; SIGformer binary interaction labels. SIGformer

Attributed Graphs

Name Type #nodes #edges #attributes #labels URL
Wiki directed 2405 17981 4973 19 [raw] [preprocessed]
Cora directed 2708 5429 1433 7 [raw] [preprocessed]
Citeseer directed 3312 4660 3703 6 [raw] [preprocessed]
Pubmed directed 19717 44338 500 3 [raw] [preprocessed]
BlogCatalog undirected 5196 343486 8189 6 [raw] [preprocessed]
PPI undirected 56944 818716 50 121 [raw] [preprocessed]
Flickr undirected 7575 479476 12047 9 [raw] [preprocessed]
Facebook undirected 4039 88234 1283 193 [raw] [preprocessed]
ArXiv undirected 169343 1157799 128 40 [raw] [preprocessed]
Reddit undirected 232,965 11,606,919 602 41 [raw] [preprocessed]
Yelp undirected 716847 6,977,410 300 100 [raw] [preprocessed]
Twitter directed 81306 1768149 216839 4065 [raw] [preprocessed]
Amazon2M undirected 2449029 61859140 100 47 [raw] [raw] [preprocessed]
Google+ directed 107614 13673453 15907 468 [raw] [preprocessed]
TWeibo directed 2320895 50655143 1657 8 [raw] [preprocessed]
MAG directed 59249719 978147253 2000 100 [raw] [preprocessed]
MAG-SC directed 10541560 265219994 2784240 8 [raw] [preprocessed]

Our datasets are also available in Pytorch-Geometric. Node attributes can be loaded as a sparse matrix using the following code

from scipy import sparse
features = sparse.load_npz("attrs.npz")

Welcome to cite our paper if you publish results based on our preprocessed datasets.

@article{yang2020scaling,
  title={Scaling Attributed Network Embedding to Massive Graphs},
  author={Yang, Renchi and Shi, Jieming and Xiao, Xiaokui and Yang, Yin and Liu, Juncheng and Bhowmick, Sourav S},
  journal={Proceedings of the VLDB Endowment},
  volume={14},
  number={1},
  pages={37--49},
  year={2021},
  publisher={VLDB Endowment}
}

Bipartite Graphs

Name |U| |V| |E| URL
Avito 27736 16589 67029 [raw] [preprocessed]
AOL 4811647 1632788 10741954 [raw] [preprocessed]
DBLP 6001 1524 29257 [raw] [preprocessed]
Movielens-1M 6040 3706 1000210 [raw] [preprocessed]
KDDCup2012 255170 1848114 2766394 [raw] [preprocessed]
Last.fm 359349 160168 17559531 [raw] [preprocessed]
Amazon-games 826767 50210 1324754 [raw] [preprocessed]
DBLP 6,001 1,308 29,256 [raw] [preprocessed]
Wikipedia 15,000 3,214 64,095 [raw] [preprocessed]
Pinterest 55,187 9,916 1,500,809 [raw] [preprocessed]
Yelp 31,668 38,048 1,561,406 [raw] [preprocessed]
MovieLens-10M 69,878 10,677 10,000,054 [raw] [preprocessed]
Last.fm 359,349 160,168 17,559,530 [raw] [preprocessed]
MIND 876,956 97,509 18,149,915 [raw] [preprocessed]
Netflix 480,189 17,770 100,480,507 [raw] [preprocessed]
Orkut 2,783,196 8,730,857 327,037,487 [raw] [preprocessed]
MAG 10,541,560 2,784,240 1,095,315,106 [raw] [preprocessed]

Welcome to cite our paper if you publish results based on our preprocessed datasets.

@inproceedings{yang2022efficient,
  title={Efficient and Effective Similarity Search over Bipartite Graphs},
  author={Yang, Renchi},
  booktitle={Proceedings of the ACM Web Conference 2022},
  pages={308--318},
  year={2022}
}

@inproceedings{yang2022scalable,
  title={Scalable and Effective Bipartite Network Embedding},
  author={Yang, Renchi and Shi, Jieming and Huang, Keke and Xiao, Xiaokui},
  booktitle={Proceedings of the 2022 International Conference on Management of Data},
  pages={1977--1991},
  year={2022}
}

Text-Attributed Graphs

Name #nodes #edges #labels / relations Domain Text location / features Task Notes Source
Cora 2,708 21,112 7 Co-citation Node & Edge Node Classification; Link Prediction Suggested fill for root README row TAGLAS
PubMed 19,717 44,338 3 Co-citation Node & Edge Node Classification; Link Prediction Suggested fill for root README row TAGLAS
ArXiv 169,343 1,166,243 40 Citation Node & Edge Node Classification Suggested fill for root README row TAGLAS
WikiCS 11,701 216,123 10 Wikipedia page Node & Edge Node Classification Suggested fill for root README row TAGLAS
Product-subset 54,025 144,638 44 used / 47 original Co-purchase Node & Edge Node Classification Suggested fill for root README row TAGLAS
FB15K237 14,541 310,116 237 relations Knowledge graph Node & Edge Link Prediction Suggested fill for root README row TAGLAS
WN18RR 40,943 93,003 11 relations Knowledge graph Node & Edge Link Prediction Suggested fill for root README row TAGLAS
MovieLens-1M 9,923 2,000,418 5 ratings Movie rating Node & Edge Link Regression / Classification Suggested fill for root README row TAGLAS
Cornell 195 304 5 Web page Raw text + PLM features Node Classification HeTGB; edge homophily 0.13 HeTGB
Texas 187 328 5 Web page Raw text + PLM features Node Classification HeTGB; edge homophily 0.12 HeTGB
Wisconsin 265 530 5 Web page Raw text + PLM features Node Classification HeTGB; edge homophily 0.20 HeTGB
Actor 4,416 12,172 5 Social Raw text + PLM features Node Classification HeTGB; edge homophily 0.56 HeTGB
Amazon 24,492 93,050 5 E-commerce Raw text + PLM features Node Classification HeTGB; edge homophily 0.38 HeTGB
ogb-products 2,449,029 61,859,140 - E-commerce Node Node Classification Survey-collected row LLMs on Graphs Survey
ogb-papers110M 111,059,956 1,615,685,872 - Academic Node Node Classification Survey-collected row LLMs on Graphs Survey
ogb-citation2 2,927,963 30,561,187 - Academic Node Link Prediction Survey-collected row LLMs on Graphs Survey
DBLP 5,259,858 36,630,661 - Academic Node Node Classification; Link Prediction Survey-collected row LLMs on Graphs Survey
MAG ~10M ~50M - Academic Node Node Classification; Link Prediction; Recommendation; Regression Survey-collected row LLMs on Graphs Survey
Goodreads-books ~2M ~20M - Books Node Node Classification; Link Prediction Survey-collected row LLMs on Graphs Survey
Amazon-items ~15.5M ~100M - E-commerce Node Node Classification; Link Prediction; Recommendation Survey-collected row LLMs on Graphs Survey
Wikidata5M ~4M ~20M - Wikipedia Node Link Prediction Survey-collected row LLMs on Graphs Survey
Twitter 176,279 2,373,956 - Social Node Node Classification; Link Prediction Survey-collected row LLMs on Graphs Survey

More in paper, paper, paper

Textual-Edge Graphs

Name #nodes #edges #labels / relations Domain Text location Task Notes Source
Goodreads-History 540,807 2,368,539 11 Book Recommendation Node & Edge Node Classification; Link Prediction TEG-DB; large scale TEG-DB
Goodreads-Crime 422,653 2,068,223 11 Book Recommendation Node & Edge Node Classification; Link Prediction TEG-DB; large scale TEG-DB
Goodreads-Children 216,624 858,586 11 Book Recommendation Node & Edge Node Classification; Link Prediction TEG-DB; large scale TEG-DB
Goodreads-Comics 148,669 631,649 11 Book Recommendation Node & Edge Node Classification; Link Prediction TEG-DB; medium scale TEG-DB
Amazon-Movie 137,411 2,724,028 399 E-commerce Node & Edge Node Classification; Link Prediction TEG-DB; medium scale TEG-DB
Amazon-Apps 31,949 62,036 62 E-commerce Node & Edge Node Classification; Link Prediction TEG-DB; small scale TEG-DB
Reddit 478,022 676,684 3 Social Networks Node & Edge Node Classification; Link Prediction TEG-DB; large scale TEG-DB
Twitter 18,761 23,764 505 Social Networks Node & Edge Node Classification; Link Prediction TEG-DB; small scale TEG-DB
Citation 169,343 1,166,243 40 Academic Node & Edge Node Classification; Link Prediction TEG-DB; large scale TEG-DB
Amazon-Movie 173,986 1,697,533 - E-commerce Edge-focused textual network Edge Classification; Link Prediction; Node Classification Edgeformers preprocessing; differs from TEG-DB Edgeformers
Amazon-Apps 100,468 752,937 - E-commerce Edge-focused textual network Edge Classification; Link Prediction; Node Classification Edgeformers preprocessing; differs from TEG-DB Edgeformers
Goodreads-Crime 385,203 1,849,236 - Book Recommendation Edge-focused textual network Edge Classification; Link Prediction Edgeformers preprocessing; also used by Link2Doc Edgeformers, Link2Doc
Goodreads-Children 192,036 734,640 - Book Recommendation Edge-focused textual network Edge Classification; Link Prediction Edgeformers preprocessing; also used by Link2Doc Edgeformers, Link2Doc
StackOverflow 129,322 281,657 - Social Edge-focused textual network Edge Classification; Link Prediction Edgeformers row; same count also appears in LLMs-on-Graphs survey Edgeformers, LLMs on Graphs Survey
Goodreads-reviews ~3M ~100M - Books Edge Edge Classification; Link Prediction Survey-collected row LLMs on Graphs Survey
Amazon-reviews ~15.5M ~200M - E-commerce Edge Edge Classification; Link Prediction Survey-collected row LLMs on Graphs Survey

Bipartite Textual-Edge Graphs

Name #source nodes #destination nodes #interactions / edges #classes Domain Text location Task Notes Source
Goodreads-Children 10,521 users 1,479 items 40,762 6 Book reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Amazon-Apps 4,390 users 5,610 items 51,073 5 E-commerce reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Amazon-Movie 4,431 users 1,819 items 61,216 5 E-commerce reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Goodreads-Crime 7,009 users 1,241 items 62,774 6 Book reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Goodreads-Poetry 47,400 users 36,412 items 154,555 6 Book reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Google-Vermont 12,655 users 5,040 items 178,168 5 Local business reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Google-Hawaii 39,215 users 10,170 items 710,948 5 Local business reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT
Amazon-Products 101,498 users 27,965 items 800,144 5 E-commerce reviews Interactions / edges Textual Interaction Classification TIN benchmark SAFT

Heterogeneous Text-Attributed Graphs

Name #nodes #edges #classes Domain Node types / text-rich types Edge types Task Notes Source
DBLP 6,343,626 50,633,439 - Academic paper text-rich; venue; author paper-paper; venue-paper; author-paper Link Prediction; Node Classification; Clustering Heterformer dataset Heterformer
Twitter 489,373 721,319 - Social media tweet and POI text-rich; hashtag; user; mention tweet-POI; user-tweet; hashtag-tweet; mention-tweet Link Prediction; Node Classification; Clustering Heterformer dataset Heterformer
Goodreads 1,373,802 42,099,586 - Books book text-rich; shelves; author; format; publisher; language code book-book; shelves-book; author-book; format-book; publisher-book; language-book Link Prediction; Node Classification; Clustering Heterformer dataset Heterformer
TMDB 24,412 104,858 4 Movie Movie; Actor; Director Movie-Actor; Movie-Director Node Classification HTAG; time split HTAG
CroVal 44,386 164,981 6 Community QA Question; User; Tag Question-Question; Question-User; Question-Tag Node Classification HTAG; time split HTAG
ArXiv 231,111 2,075,692 40 Academic Paper; Author; Field of Study Paper-Paper; Paper-Author; Paper-FoS Node Classification HTAG; time split HTAG
Book 786,257 9,035,291 8 Literature Book; Author; Publisher Book-Book; Book-Author; Book-Publisher Node Classification HTAG; time split HTAG
DBLP 1,989,010 29,830,033 9 Academic Paper; Author; Field of Study Paper-Paper; Paper-Author; Paper-FoS Node Classification HTAG; time split; different from Heterformer DBLP HTAG
Patent 5,646,139 8,833,738 120 Patent Patent; Inventor; Examiner Patent-Inventor; Patent-Examiner Node Classification HTAG; time split HTAG
CITE 438,304 1,220,373 85 Catalytic materials Paper; Author; Keywords; Journal Paper-Paper; Paper-Author; Paper-Keywords; Paper-Journal Node Classification Catalyst-materials citation graph CITE

Multiplex Text-Attributed Graphs

Name #nodes Relation edge counts Domain Task Notes Source
Geology 431,834 cb: 1,000,000; sa: 1,000,000; sv: 1,000,000; cr: 1,000,000; ccb: 1,000,000 Academic Multiplex representation learning METAG benchmark METAG
Mathematics 490,551 cb: 1,000,000; sa: 1,000,000; sv: 1,000,000; cr: 1,000,000; ccb: 1,000,000 Academic Multiplex representation learning METAG benchmark METAG
Clothes 208,318 cop: 100,000; cov: 100,000; bt: 100,000; cob: 50,000 E-commerce Multiplex representation learning METAG benchmark METAG
Home 192,150 cop: 100,000; cov: 100,000; bt: 50,000; cob: 100,000 E-commerce Multiplex representation learning METAG benchmark METAG
Sports 189,526 cop: 100,000; cov: 100,000; bt: 50,000; cob: 100,000 E-commerce Multiplex representation learning METAG benchmark METAG

Text-Attributed Hypergraphs

Name #nodes #hyperedges #classes Domain Hyperedge construction / avg size Avg. tokens Task Notes Source
Cora 1,434 1,579 7 Academic Co-citation - Node Classification HyperBERT preprocessing HyperBERT
PubMed 3,840 7,963 3 Academic Co-citation - Node Classification HyperBERT preprocessing HyperBERT
DBLP-A 2,591 2,690 6 Academic Co-authorship - Node Classification HyperBERT preprocessing HyperBERT
Cora-CA 2,388 1,072 7 Academic Co-authorship - Node Classification HyperBERT preprocessing HyperBERT
IMDB 3,939 2,015 3 Movie Movie-actor - Node Classification HyperBERT preprocessing HyperBERT
Citeseer 1,778 2,118 6 Academic Avg. size 2 198 Node Classification HiTeC preprocessing HiTeC
Cora 2,708 1,579 7 Academic Avg. size 3 189 Node Classification HiTeC preprocessing; different from HyperBERT Cora HiTeC
History 41,551 169,454 12 E-commerce Avg. size 9 300 Node Classification HiTeC preprocessing HiTeC
Photo 48,362 212,247 12 E-commerce Avg. size 11 189 Node Classification HiTeC preprocessing HiTeC
Computers 87,229 277,539 10 E-commerce Avg. size 9 115 Node Classification HiTeC preprocessing HiTeC
Fitness 173,055 1,468,229 13 E-commerce Avg. size 11 28 Node Classification HiTeC preprocessing HiTeC

Dynamic Text-Attributed Graphs

For bipartite GDGB rows, #nodes is reported as source/destination node counts.

Name #nodes #edges Edge labels / categories Timestamps Domain Text attributes Bipartite Task / benchmark Source
Enron 42,711 797,907 10 1,006 E-mail Node & Edge No DTGB discriminative DyTAG benchmark DTGB
GDELT 6,786 1,339,245 237 2,591 Knowledge graph Node & Edge No DTGB discriminative DyTAG benchmark DTGB
ICEWS1819 31,796 1,100,071 266 730 Knowledge graph Node & Edge No DTGB discriminative DyTAG benchmark DTGB
Stack elec 397,702 1,262,225 2 5,224 Multi-round dialogue Node & Edge Yes DTGB discriminative DyTAG benchmark DTGB
Stack ubuntu 674,248 1,497,006 2 4,972 Multi-round dialogue Node & Edge Yes DTGB discriminative DyTAG benchmark DTGB
Googlemap CT 111,168 1,380,623 5 55,521 E-commerce Node & Edge Yes DTGB discriminative DyTAG benchmark DTGB
Amazon movies 293,566 3,217,324 5 7,287 E-commerce Node & Edge Yes DTGB discriminative DyTAG benchmark DTGB
Yelp 2,138,242 6,990,189 5 6,036 E-commerce Node & Edge Yes DTGB discriminative DyTAG benchmark DTGB
Sephora 210,357 / 2,274 801,234 5 5,314 E-commerce Node & Edge Yes GDGB generative DyTAG benchmark GDGB
Dianping 158,541 / 88,118 1,990,409 5 745,151 E-commerce Node & Edge Yes GDGB generative DyTAG benchmark GDGB
WikiRevision 75,622 / 3,204 2,778,732 2 2,766,153 Web Interaction Node & Edge Yes GDGB generative DyTAG benchmark GDGB
WikiLife 406,148 / 54,513 1,996,520 24 1,810 Celebrity Biography Node & Edge Yes GDGB generative DyTAG benchmark GDGB
IMDB 125,714 1,534,162 20 122 Movie Collaboration Node & Edge No GDGB generative DyTAG benchmark GDGB
WeiboTech 20,767 109,345 2 79,925 Social Network Node & Edge No GDGB generative DyTAG benchmark GDGB
WeiboDaily 66,500 354,098 2 293,662 Social Network Node & Edge No GDGB generative DyTAG benchmark GDGB
Cora 48,797 110,788 5 8,274 Citation Node & Edge No GDGB generative DyTAG benchmark GDGB

Multi-Modal Graphs

Name Graph size / scale #nodes / entities / subjects #edges / interactions / triples #labels / relations Modality Domain Task Notes Source
MM-CoDEx-s 14,298 train+; 784 valid+; 802 test+ triples 1,383 entities 15,884 positive triples 39 relations Text, Vision Multi-modal KGs Knowledge Graph Completion Missing KGC row from root MM-Graph table MM-Graph
MM-CoDEx-m 47,617 train+; 2,628 valid+; 2,595 test+ triples 7,697 entities 52,840 positive triples 51 relations Text, Vision Multi-modal KGs Knowledge Graph Completion Missing KGC row from root MM-Graph table MM-Graph
Tiktok 726,065 interactions 76,085 items; 36,656 users 726,065 interactions - Visual 128; Acoustic 128; Text 128 Micro-video recommendation Recommendation MMGCN evaluation dataset MMGCN
Kwai 1,664,305 interactions 329,510 items; 22,611 users 1,664,305 interactions - Visual 2,048; Text 128 Micro-video recommendation Recommendation MMGCN evaluation dataset MMGCN
MovieLens 1,239,508 interactions 5,986 items; 55,485 users 1,239,508 interactions - Visual 2,048; Acoustic 128; Text 100 Movie recommendation Recommendation MMGCN evaluation dataset MMGCN
ADNI 417 subjects; 17 non-image features 417 subjects - 3 classes 3D medical images + tabular features Brain / Alzheimer Medical classification HetMed dataset HetMed
OASIS-3 979 subjects; 19 non-image features 979 subjects - 4 classes 3D medical images + tabular features Brain / Alzheimer Medical classification HetMed dataset HetMed
ABIDE 977 subjects; 14 non-image features 977 subjects - 2 classes 3D medical images + tabular features Brain / Autism Medical classification HetMed dataset HetMed
Duke-Breast 614 subjects; 25 non-image features 614 subjects - 3 classes Medical images + tabular features Breast / Tumor Medical classification HetMed dataset HetMed
CMMD 1,774 subjects; 4 non-image features 1,774 subjects - 2 classes Medical images + tabular features Breast / Tumor Medical classification HetMed dataset HetMed
G2MF-Urban 100K nodes; 2M edges 100K nodes 2M edges - Text + Vision Urban planning Node Classification Survey-collected row MG-LLM Survey
Pan-Cancer Atlas 11K samples from 33 cancer types 11K samples - 33 cancer types Multi-omics Biomedical repository Node Classification Survey-collected row MG-LLM Survey
TCGA-BRCA 1,084 breast tumor samples 1,084 samples - - Multi-omics Biomedical repository Node Classification Survey-collected row MG-LLM Survey
OMG-NAS Tencent 8K nodes; 60K edges 8K nodes 60K edges - Text + Vision Website articles Node Classification Survey-collected row MG-LLM Survey
OMG-NAS Amazon 100K nodes; 300K edges 100K nodes 300K edges - Text + Vision E-commerce products Node Classification Survey-collected row MG-LLM Survey
VTKG-I&C 130 entities; 842 triples 130 entities 842 triples - Text + Vision Multi-modal KGs Link Prediction Survey-collected row MG-LLM Survey
TIVA-KG 50K entities; 200K triples 50K entities 200K triples - Text + Vision + Audio Multi-modal KGs Link Prediction Survey-collected row MG-LLM Survey
OMG-NAS Recipe 20K nodes; 160K edges 20K nodes 160K edges - Text + Vision Food recipes Graph Classification Survey-collected row MG-LLM Survey
Large-RG 500K nodes 500K nodes - - Text + Vision Food recipes Graph Classification Survey-collected row MG-LLM Survey
GQA 113K images; 23M questions 113K images - - Text + Vision Scene graphs Graph Question Answering Survey-collected row MG-LLM Survey
CLEVR 100K images; 1M questions 100K images - - Text + Vision Scene graphs Graph Question Answering Survey-collected row MG-LLM Survey
SceneGraph-VQA 50K images 50K images - - Text + Vision Scene graphs Graph Question Answering Survey-collected row MG-LLM Survey
MARS & MarKG 34K triples; 13K questions - 34K triples - Text + Vision Multi-modal KGs Graph Reasoning Survey-collected row MG-LLM Survey
Richpedia 3M entities 3M entities - - Text + Vision Multi-modal KGs Text Generation Survey-collected row MG-LLM Survey
ART500K 311K nodes; 643M edges 311K nodes 643M edges - Text + Vision Artwork relationships Image Generation Survey-collected row MG-LLM Survey
Amazon Coview 178K nodes; 3M edges 178K nodes 3M edges - Text + Vision E-commerce products Image Generation Survey-collected row MG-LLM Survey
Goodreads 93K nodes; 637K edges 93K nodes 637K edges - Text + Vision Book recommendation Image Generation Survey-collected row MG-LLM Survey

Graph-level Datasets

Name #graphs #classes / tasks Avg. #nodes Avg. #edges Domain Task / notes Source
Chemblpre 365,065 - 25.87 55.92 Molecular Graph-level property prediction TAGLAS
molproperties 363,336 - 25.57 55.32 Molecular Graph Question Answering TAGLAS
PCBA 437,929 - 25.97 56.20 Molecular Graph-level property prediction TAGLAS
HIV 41,127 - 25.51 54.94 Molecular Graph-level property prediction TAGLAS
BBBP 2,039 - 24.06 51.91 Molecular Graph-level property prediction TAGLAS
BACE 1,513 - 34.09 73.72 Molecular Graph-level property prediction TAGLAS
toxcast 8,575 - 18.76 38.50 Molecular Graph-level property prediction TAGLAS
esol 1,128 - 13.29 27.35 Molecular Graph-level property prediction TAGLAS
freesolv 642 - 8.72 16.76 Molecular Graph-level property prediction TAGLAS
lipo 4,200 - 27.04 59.00 Molecular Graph-level property prediction TAGLAS
cyp450 16,896 - 24.52 53.02 Molecular Graph-level property prediction TAGLAS
tox21 7,831 - 18.57 38.59 Molecular Graph-level property prediction TAGLAS
muv 93,087 - 24.23 52.56 Molecular Graph-level property prediction TAGLAS
ExplaGraphs 2,766 - 5.17 4.25 Commonsense Graph Question Answering TAGLAS
SceneGraphs 100,000 - 19.13 68.44 Scene graph Graph Question Answering TAGLAS
MAG240M-subset 1 153 classes 5,875,010 26,434,726 Citation Node Classification on one large graph TAGLAS
Ultrachat200k 449,929 - 3.72 2.72 Conversation Graph Question Answering TAGLAS
Wikikg90m 1 - 91,230,610 1,202,155,622 Knowledge graph Link Prediction on one large graph TAGLAS
MUTAG 188 2 17.93 19.79 Small molecules Graph Classification TUDataset
NCI1 4,110 2 29.87 32.30 Small molecules Graph Classification TUDataset
Mutagenicity 4,337 2 30.32 30.77 Small molecules Graph Classification TUDataset
PROTEINS 1,113 2 39.06 72.82 Bioinformatics Graph Classification TUDataset
ENZYMES 600 6 32.63 62.14 Bioinformatics Graph Classification TUDataset
DD 1,178 2 284.32 715.66 Bioinformatics Graph Classification TUDataset
COIL-RAG 3,900 100 3.01 3.02 Computer vision Graph Classification TUDataset
MSRC_21 563 20 77.52 198.32 Computer vision Graph Classification TUDataset
COLLAB 5,000 3 74.49 2,457.78 Social networks Graph Classification TUDataset
IMDB-BINARY 1,000 2 19.77 96.53 Social networks Graph Classification TUDataset
IMDB-MULTI 1,500 3 13.00 65.94 Social networks Graph Classification TUDataset
REDDIT-BINARY 2,000 2 429.63 497.75 Social networks Graph Classification TUDataset
REDDIT-MULTI-5K 4,999 5 508.52 594.87 Social networks Graph Classification TUDataset

Dataset Repositories

Name Type Collected by
SNAP Graphs & Networks Stanford
LAW Graphs & Networks UNIMI
BioSNAP Biomedical Networks Stanford
KONECT Graphs & Networks Jérôme Kunegis
Aminer Academic Networks AMiner
UCI Network Data Repository Graphs & Networks UCI Datalab
Network Repository Graphs & Networks -
Open Academic Graph Academic Networks Microsoft
Open Graph Benchmark Graphs & Networks Stanford
TuDatasets Graphs & Networks Christopher Morris, etc.
StreamingGraphs Streaming Graphs Yibo Yao
ARB Graphs & Networks Austin R. Benson
SuiteSparse Matrix Collection Matrix/Graphs TAMU
Web Data Commons Hyperlink Graphs/Web Tables/RDFa University of Mannheim
Yahoo Webscope Datasets Graphs/Ratings/Languages/Advertising Yahoo
UCI Machine Learning Repository Multivariate/Text/Time-Series UCI
Yelp Open Dataset businesses/reviews/user data Yelp
Recommender Systems Datasets graphs/interactions/reviews/ratings UCSD
MIcrosoft News Dataset user behavior logs Microsoft
Search Query Logs query logs Jeff Huang
AOL DS query logs Ricardo Campos
AWS - Amazon
Kaggle Datasets - Kaggle
OpenML - OpenML
Datasets - -
Netzschleuder - -

About

A curated list of graph datasets of various types, including plaingraphs, attributed graphs, bipartite graphs, text-attributed graphs, multi-modal graphs, temporal graphs, etc.

Topics

bipartite-graphsdynamic-graphgraphsknowledge-graphmulti-modal-graphtemporal-graphtext-attributed-graph

Resources

Readme
MIT license
Activity
Custom properties

Stars

10 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

Packages

Contributors