Leaderboards for Link Property Prediction
Check leaderboards for
- ogbl-ppa
- ogbl-collab
- ogbl-ddi
- ogbl-citation2
- ogbl-wikikg2
- ogbl-biokg
- ogbl-vessel
The bold method name indicates that the implementation is official (by the author of the original paper).
Package denotes the required package version for each dataset to be eligible for the leaderboard.
Leaderboard for ogbl-ppa
The Hits@100 score on the test and validation sets. The higher, the better.
Package: >=1.1.1
| Rank | Method | Ext. data | Test Hits@100 | Validation Hits@100 | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | GraphGPT(SMTP) | No | 0.7655 ± 0.0067 | 0.7712 ± 0.0065 | Qifang Zhao(Alibaba) | Paper, Code | 2,032,859,200 | 32 Nvidia L40S | Oct 28, 2024 |
| 2 | Refined-GAE | No | 0.7374 ± 0.0092 | 0.7390 ± 0.0180 | Weishuo Ma(Peking University) | Paper, Code | 295,848,449 | RTX 4090 (24GB) | Nov 23, 2024 |
| 3 | Refined-GAE | No | 0.7334 ± 0.0092 | 0.7391 ± 0.0178 | Weishuo Ma(Peking University) | Paper, Code | 295,848,449 | RTX 4090 (24GB) | Nov 7, 2024 |
| 4 | GraphGPT(SMTP) | No | 0.6876 ± 0.0067 | 0.7017 ± 0.0044 | Qifang Zhao(Alibaba) | Paper, Code | 145,263,360 | 8 Nvidia L40S | Oct 8, 2024 |
| 5 | MPLP | No | 0.6524 ± 0.0150 | 0.6685 ± 0.0073 | Kaiwen Dong (University of Notre Dame) | Paper, Code | 147,794,531 | NVIDIA A100 GPU (80G RAM) | Jan 16, 2024 |
| 6 | GCN (node embedding) | No | 0.6354 ± 0.0121 | 0.6524 ± 0.0096 | Shuming Liang (UTS) | Paper, Code | 148,144,898 | NVIDIA A100 (40G) | Nov 24, 2024 |
| 7 | SIEG | No | 0.6322 ± 0.0174 | 0.6533 ± 0.0234 | anonymous20221001 | Paper, Code | 1,993,965 | P100 (16GB) | Mar 24, 2023 |
| 8 | **Neural Common Neighbor ** | No | 0.6119 ± 0.0085 | 0.6021 ± 0.0037 | Xiyuan Wang (Peking University) | Paper, Code | 33,538 | RTX 4090 (24GB) | Mar 17, 2023 |
| 9 | NGNN + SEAL | No | 0.5971 ± 0.0245 | 0.5995 ± 0.0205 | Yakun Song(DGL) | Paper, Code | 735,426 | Tesla T4(16GB) | Sep 21, 2022 |
| 10 | SUREL+ | No | 0.5432 ± 0.0044 | 0.5492 ± 0.0112 | Haoteng Yin (Purdue University) | Paper, Code | 52,802 | NVIDIA A100 (80G) | Feb 20, 2023 |
| 11 | RA+Edge Proposal Set | No | 0.5324 ± 0.0000 | 0.5142 ± 0.0000 | Qian Huang (Stanford) | Paper, Code | 0 | GeForce RTX 2080 Ti (11GB GPU) | Oct 5, 2021 |
| 12 | MLP+CN&RA&AA | No | 0.5062 ± 0.0035 | 0.4906 ± 0.0029 | Shitao Lu | Paper, Code | 163,330 | Geforce GTX 1080 Ti (11GB) | Aug 9, 2021 |
| 13 | BUDDY | No | 0.4934 ± 0.0098 | 0.4931 ± 0.0025 | Ben Chamberlain | Paper, Code | 644 | Tesla K80 (12GB GPU) | Aug 23, 2023 |
| 14 | Resource Allocation | No | 0.4933 ± 0.0000 | 0.4722 ± 0.0000 | Shen Fan(Alibaba Group) | Paper, Code | 0 | Tesla-V100(32GB GPU) | Jul 5, 2021 |
| 15 | SEAL | No | 0.4880 ± 0.0316 | 0.5125 ± 0.0252 | Muhan Zhang | Paper, Code | 709,122 | GeForce RTX 2080S (8GB GPU) | Oct 14, 2020 |
| 16 | S3GRL (PoS Plus) | No | 0.4242 ± 0.0180 | 0.6512 ± 0.0109 | Paul Louis (Ontario Tech University) | Paper, Code | 32,270,001 | GeForce GTX 1080 Ti (11 GB) | May 2, 2023 |
| 17 | AGDN | No | 0.4123 ± 0.0159 | 0.4332 ± 0.0092 | Chuxiong Sun | Paper, Code | 36,904,259 | Tesla V100 (16GB GPU) | Sep 2, 2022 |
| 18 | NGNN + GraphSAGE | No | 0.4005 ± 0.0138 | 0.4058 ± 0.0123 | Yakun Song(DGL) | Paper, Code | 556,033 | Tesla-V100(16GB GPU) | Aug 16, 2022 |
| 19 | NGNN + GCN | No | 0.3683 ± 0.0099 | 0.3834 ± 0.0082 | Yakun Song(DGL) | Paper, Code | 410,113 | Tesla-V100(16GB GPU) | Aug 16, 2022 |
| 20 | Adamic Adar | No | 0.3245 ± 0.0000 | 0.3268 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 21 | Matrix Factorization | No | 0.3229 ± 0.0094 | 0.3228 ± 0.0428 | Matthias Fey – OGB team | Paper, Code | 147,662,849 | GeForce RTX 2080 (11GB GPU) | May 1, 2020 |
| 22 | DeepWalk | No | 0.2888 ± 0.0153 | Please tell us | Hao Xiong (DGL) | Paper, Code | 150,138,741 | g4dn.2xlarge, T4 (15GB GPU) | Jul 23, 2020 |
| 23 | Common Neighbor | No | 0.2765 ± 0.0000 | 0.2823 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 24 | DeepWalk | No | 0.2302 ± 0.0163 | Please tell us | Hao Xiong (DGL) | Paper, Code | 150,138,741 | g4dn.2xlarge, T4 (15GB GPU) | Jun 30, 2020 |
| 25 | Node2vec | No | 0.2226 ± 0.0083 | 0.2253 ± 0.0088 | Matthias Fey – OGB team | Paper, Code | 73,878,913 | GeForce RTX 2080 (11GB GPU) | May 1, 2020 |
| 26 | GCN | No | 0.1867 ± 0.0132 | 0.1845 ± 0.0140 | Matthias Fey – OGB team | Paper, Code | 278,529 | GeForce RTX 2080 (11GB GPU) | Jun 25, 2020 |
| 27 | GraphSAGE | No | 0.1655 ± 0.0240 | 0.1724 ± 0.0264 | Matthias Fey – OGB team | Paper, Code | 424,449 | GeForce RTX 2080 (11GB GPU) | Jun 25, 2020 |
Leaderboard for ogbl-collab
The Hits@50 score on the test and validation sets. The higher, the better.
Package: >=1.2.1
| Rank | Method | Ext. data | Test Hits@50 | Validation Hits@50 | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | HyperFusion | No | 0.7129 ± 0.0018 | 0.7385 ± 0.0099 | Xinwei Zhang (Tsinghua University) | Paper, Code | 1,064,446,212 | RTX 3080 | Feb 24, 2024 |
| 2 | GIDN@YITU | No | 0.7096 ± 0.0055 | 0.9620 ± 0.0040 | Zixiao Wang, Yu Zhang(ZhejiangLab, HUST) | Paper, Code | 60,449,025 | DepGraph@SCTS/CGCL | Oct 10, 2022 |
| 3 | PLNLP + SIGN | No | 0.7087 ± 0.0033 | 1.0000 ± 0.0000 | Liang Yao (Tencent) | Paper, Code | 34,980,864 | Tesla-P40 (24G GPU) | Apr 7, 2022 |
| 4 | PLNLP (random walk aug.) | No | 0.7059 ± 0.0029 | 1.0000 ± 0.0000 | Zhitao Wang (WeChat@Tencent) | Paper, Code | 34,980,864 | Tesla-P40 (24G GPU) | Dec 21, 2021 |
| 5 | HOP-REC | No | 0.7012 ± 0.0016 | 1.0000 ± 0.0000 | Bo-Yu Lin (CFDA & CLIP Labs) | Paper, Code | 30,191,104 | CPU | Oct 21, 2021 |
| 6 | PLNLP+ LRGA | No | 0.6909 ± 0.0055 | 1.0000 ± 0.0000 | Hao Xu | Paper, Code | 35,200,656 | NVIDIA Tesla V100 (32GB GPU) | Jun 26, 2022 |
| 7 | PLNLP (val as input) | No | 0.6872 ± 0.0052 | 1.0000 ± 0.0000 | Zhitao Wang (WeChat@Tencent) | Paper, Code | 35,112,192 | Tesla-P40 (24G GPU) | Dec 7, 2021 |
| 8 | Refined-GAE | No | 0.6816 ± 0.0041 | 1.0000 ± 0.0000 | Weishuo Ma(Peking University) | Paper, Code | 13,816,833 | RTX 4090 (24GB) | Nov 7, 2024 |
| 8 | Refined-GAE | No | 0.6816 ± 0.0041 | 1.0000 ± 0.0000 | Weishuo Ma(Peking University) | Paper, Code | 126,669,825 | RTX 4090 (24GB) | Nov 23, 2024 |
| 9 | TopoLink | No | 0.6792 ± 0.0074 | 0.6771 ± 0.0083 | Lizhi Liu (China UnionPay) | Paper, Code | 483,363,845 | RTX 4090 (24GB GPU) | Sep 9, 2024 |
| 10 | S3GRL (PoS Plus) | No | 0.6683 ± 0.0030 | 0.9861 ± 0.0006 | Paul Louis (Ontario Tech University) | Paper, Code | 5,913,025 | GeForce GTX 1080 Ti (11 GB) | Apr 25, 2023 |
| 11 | ELPH | No | 0.6636 ± 0.5876 | 0.6631 ± 0.0021 | Ben Chamberlain | Paper, Code | 3,284,065 | Tesla K80 (12GB GPU) | Jul 3, 2023 |
| 12 | BUDDY | No | 0.6572 ± 0.0053 | 0.6621 ± 0.0016 | Ben Chamberlain | Paper, Code | 1,184,867 | Tesla K80 (12GB GPU) | Jul 2, 2023 |
| 13 | Adamic Adar+Edge Proposal Set | No | 0.6548 ± 0.0000 | 0.9735 ± 0.0000 | Qian Huang (Cornell University) | Paper, Code | 0 | GeForce RTX 2080 Ti (11GB GPU) | Jul 13, 2021 |
| 14 | SEAL-nofeat (val as input) | No | 0.6474 ± 0.0043 | 0.6495 ± 0.0043 | Muhan Zhang | Paper, Code | 501,570 | NVIDIA Tesla V100 (32GB GPU) | Jun 16, 2021 |
| 15 | Adamic Adar | No | 0.6417 ± 0.0000 | 0.6349 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 16 | Common Neighbor | No | 0.6137 ± 0.0000 | 0.6036 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 17 | SEAL-nofeat | No | 0.5471 ± 0.0049 | 0.6495 ± 0.0043 | Muhan Zhang | Paper, Code | 501,570 | NVIDIA Tesla V100 (32GB GPU) | Jun 16, 2021 |
| 18 | GraphSAGE (val as input) | No | 0.5463 ± 0.0112 | 0.5688 ± 0.0077 | Matthias Fey – OGB team | Paper, Code | 460,289 | GeForce RTX 2080 (11GB GPU) | Oct 19, 2020 |
| 19 | NGNN + GraphSAGE | No | 0.5359 ± 0.0056 | 0.6281 ± 0.0046 | Yakun Song(DGL) | Paper, Code | 591,873 | Tesla-V100(16GB GPU) | Aug 16, 2022 |
| 20 | NGNN + GCN | No | 0.5348 ± 0.0040 | 0.6273 ± 0.0040 | Yakun Song(DGL) | Paper, Code | 428,033 | Tesla-V100(16GB GPU) | Aug 16, 2022 |
| 21 | DeeperGCN | No | 0.5273 ± 0.0047 | 0.6187 ± 0.0045 | Guohao Li - DeepGCNs.org | Paper, Code | 117,383 | NVIDIA Tesla V100 (32GB GPU) | Oct 21, 2020 |
| 22 | LRGA + GCN | No | 0.5221 ± 0.0072 | 0.6088 ± 0.0059 | Omri Puny | Paper, Code | 1,069,489 | Tesla P100 (16GB GPU) | Aug 5, 2020 |
| 23 | DeepWalk | No | 0.5037 ± 0.0034 | Please tell us | Hao Xiong (DGL) | Paper, Code | 61,390,187 | g4dn.2xlarge, T4 (15GB GPU) | Jun 30, 2020 |
| 24 | Node2vec | No | 0.4888 ± 0.0054 | 0.5703 ± 0.0052 | Matthias Fey – OGB team | Paper, Code | 30,322,945 | GeForce RTX 2080 (11GB GPU) | Jun 22, 2020 |
| 25 | GraphSAGE | No | 0.4810 ± 0.0081 | 0.5688 ± 0.0077 | Matthias Fey – OGB team | Paper, Code | 460,289 | GeForce RTX 2080 (11GB GPU) | Jun 24, 2020 |
| 26 | GCN (val as input) | No | 0.4714 ± 0.0145 | 0.5263 ± 0.0115 | Matthias Fey – OGB team | Paper, Code | 296,449 | GeForce RTX 2080 (11GB GPU) | Oct 19, 2020 |
| 27 | GCN | No | 0.4475 ± 0.0107 | 0.5263 ± 0.0115 | Matthias Fey – OGB team | Paper, Code | 296,449 | GeForce RTX 2080 (11GB GPU) | Jun 24, 2020 |
| 28 | Matrix Factorization | No | 0.3886 ± 0.0029 | 0.4896 ± 0.0029 | Matthias Fey – OGB team | Paper, Code | 60,514,049 | GeForce RTX 2080 (11GB GPU) | Jun 22, 2020 |
Leaderboard for ogbl-ddi
The Hits@20 score on the test and validation sets. The higher, the better.
Package: >=1.2.1
| Rank | Method | Ext. data | Test Hits@20 | Validation Hits@20 | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | HyperFusion | No | 0.9972 ± 0.0004 | 0.9956 ± 0.0001 | Xinwei Zhang(Tsinghua University) | Paper, Code | 976,022,023 | RTX 3080 | Feb 24, 2024 |
| 2 | ELGNN | No | 0.9777 ± 0.0037 | 0.8965 ± 0.0021 | Zhen Hao Wong (EE Tsinghua) | Paper, Code | 10,512,391 | NVIDIA A100 GPU (80G RAM) | Oct 23, 2023 |
| 3 | GCN (node embedding) | No | 0.9549 ± 0.0073 | 0.9098 ± 0.0294 | Shuming Liang (UTS) | Paper, Code | 5,125,250 | NVIDIA Quadro P6000 (24G) | Nov 24, 2024 |
| 4 | GIDN@YITU | No | 0.9542 ± 0.0000 | 0.8258 ± 0.0000 | Zixiao Wang, Yu Zhang(Zhejiang Lab,HUST) | Paper, Code | 3,506,691 | DepGraph@SCTS/CGCL | Oct 11, 2022 |
| 5 | AGDN (AUC loss) | No | 0.9538 ± 0.0094 | 0.8943 ± 0.0281 | Chuxiong Sun | Paper, Code | 3,506,691 | Tesla V100 (16GB GPU) | Sep 2, 2022 |
| 6 | Refined-GAE | No | 0.9443 ± 0.0057 | 0.7526 ± 0.0240 | Weishuo Ma(Peking University) | Paper, Code | 13,816,833 | RTX 4090 (24GB) | Nov 7, 2024 |
| 6 | Refined-GAE | No | 0.9443 ± 0.0057 | 0.7979 ± 0.0159 | Weishuo Ma(Peking University) | Paper, Code | 13,816,833 | RTX 4090 (24GB) | Nov 23, 2024 |
| 7 | PSG | No | 0.9284 ± 0.0047 | 0.8306 ± 0.0134 | Jingsong Lv(Zhuque@ZhejiangLab) | Paper, Code | 3,499,009 | Tesla V100(32G GPU) | Aug 12, 2022 |
| 8 | PLNLP | No | 0.9088 ± 0.0313 | 0.8242 ± 0.0253 | Zhitao Wang (WeChat@Tencent) | Paper, Code | 3,497,473 | Tesla-P40(24GB GPU) | Dec 7, 2021 |
| 9 | GDNN | No | 0.9037 ± 0.0193 | 0.8599 ± 0.0286 | Haifan Zhou(Ainnocence) | Paper, Code | 3,761,665 | A40-NVLink(48GB) | Aug 30, 2022 |
| 10 | GraphSAGE + Edge Attr | No | 0.8781 ± 0.0474 | 0.8044 ± 0.0404 | Jing Yang | Paper, Code | 3,761,665 | Tesla V100 (32GB) | Aug 9, 2021 |
| 11 | CFLP (w/ JKNet) | No | 0.8608 ± 0.0198 | 0.8405 ± 0.0284 | Tong Zhao | Paper, Code | 837,635 | GeForce RTX 2080 Ti (11GB GPU) | Nov 17, 2021 |
| 11 | CFLP (w/ JKNet) | No | 0.8608 ± 0.0198 | 0.8405 ± 0.0284 | Tong Zhao (University of Notre Dame) | Paper, Code | 837,635 | GeForce RTX 2080 Ti (11GB GPU) | Nov 17, 2021 |
| 12 | GraphSAGE+anchor distance | No | 0.8239 ± 0.0437 | 0.8206 ± 0.0298 | Boning Li (SJTU) | Paper, Code | 3,760,134 | Tesla V100 (16GB GPU) | May 20, 2021 |
| 13 | NeuralCommonNeighbor | No | 0.8232 ± 0.0610 | 0.7172 ± 0.0025 | Xiyuan Wang (Peking University) | Paper, Code | 1,412,098 | RTX 4090 (24GB) | Mar 17, 2023 |
| 14 | ELPH | No | 0.7704 ± 0.0582 | 0.6928 ± 0.0096 | Ben Chamberlain | Paper, Code | 2,910,817 | Tesla K80 (12GB GPU) | Jul 5, 2023 |
| 15 | DEA + JKNet | No | 0.7672 ± 0.0265 | 0.6713 ± 0.0071 | Yichen Yang | Paper, Code | 1,763,329 | Tesla T4 (16GB GPU) | Mar 21, 2021 |
| 16 | BUDDY | No | 0.7654 ± 0.0459 | 0.6927 ± 0.0054 | Ben Chamberlain | Paper, Code | 2,712,931 | Tesla K80 (12GB GPU) | Jul 4, 2023 |
| 17 | GraphSAGE+Edge Proposal Set | No | 0.7495 ± 0.0317 | 0.6696 ± 0.0198 | Qian Huang (Cornell University) | Paper, Code | 1,421,057 | GeForce RTX 2080 Ti (11GB GPU) | Jul 21, 2021 |
| 18 | LRGA+GCN(Node2Vec+Augment) | No | 0.7385 ± 0.0871 | 0.7225 ± 0.0047 | Chuanqi Chen | Paper, Code | 10,235,281 | Tesla V100 (32GB) | Mar 21, 2021 |
| 19 | MAD Learning | No | 0.6781 ± 0.0294 | 0.7010 ± 0.0082 | Yi Luo | Paper, Code | 1,228,897 | Geforce GTX 1080 Ti (11GB GPU) | Feb 10, 2021 |
| 20 | LRGA + GCN | No | 0.6230 ± 0.0912 | 0.6675 ± 0.0058 | Omri Puny | Paper, Code | 1,576,081 | Tesla P100 (16GB GPU) | Aug 5, 2020 |
| 21 | GCN+JKNet | No | 0.6056 ± 0.0869 | 0.6776 ± 0.0095 | Horace He (Cornell) | Paper, Code | 1,421,571 | Geforce GTX 1080 Ti (11GB GPU) | Sep 15, 2020 |
| 22 | NGNN + GraphSAGE | No | 0.5770 ± 0.1523 | 0.7323 ± 0.0040 | Yakun Song(DGL) | Paper, Code | 1,618,433 | Tesla-V100(16GB GPU) | Aug 16, 2022 |
| 23 | NGNN + GCN | No | 0.5483 ± 0.1581 | 0.7121 ± 0.0038 | Yakun Song(DGL) | Paper, Code | 1,487,361 | Tesla-V100(16GB GPU) | Aug 16, 2022 |
| 24 | GraphSAGE | No | 0.5390 ± 0.0474 | 0.6262 ± 0.0037 | Matthias Fey – OGB team | Paper, Code | 1,421,057 | GeForce RTX 2080 (11GB GPU) | Jun 25, 2020 |
| 25 | GCN | No | 0.3707 ± 0.0507 | 0.5550 ± 0.0208 | Matthias Fey – OGB team | Paper, Code | 1,289,985 | GeForce RTX 2080 (11GB GPU) | Jun 25, 2020 |
| 26 | SEAL | No | 0.3056 ± 0.0386 | 0.2849 ± 0.0269 | Muhan Zhang | Paper, Code | 531,138 | GeForce RTX 2080S (8GB GPU) | Oct 14, 2020 |
| 27 | DeepWalk | No | 0.2642 ± 0.0610 | Please tell us | Hao Xiong (DGL) | Paper, Code | 11,543,913 | g4dn.2xlarge, T4 (15GB GPU) | Jul 23, 2020 |
| 28 | Node2vec | No | 0.2326 ± 0.0209 | 0.3292 ± 0.0121 | Matthias Fey – OGB team | Paper, Code | 645,249 | GeForce RTX 2080 (11GB GPU) | Jun 22, 2020 |
| 29 | DeepWalk | No | 0.2246 ± 0.0290 | Please tell us | Hao Xiong (DGL) | Paper, Code | 1,543,913 | g4dn.2xlarge, T4 (15GB GPU) | Jun 30, 2020 |
| 30 | Adamic Adar | No | 0.1861 ± 0.0000 | 0.0966 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 31 | Common Neighbor | No | 0.1773 ± 0.0000 | 0.0947 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 32 | Matrix Factorization | No | 0.1368 ± 0.0475 | 0.3370 ± 0.0264 | Matthias Fey – OGB team | Paper, Code | 1,224,193 | GeForce RTX 2080 (11GB GPU) | Jun 10, 2020 |
Leaderboard for ogbl-citation2
The MRR score on the test and validation sets. The higher, the better.
Package: >=1.2.4
Deprecated ogbl-citation leaderboard can be found here.
| Rank | Method | Ext. data | Test MRR | Validation MRR | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | GraphGPT(d1n30) | No | 0.9305 ± 0.0020 | 0.9295 ± 0.0022 | Qifang Zhao(Alibaba) | Paper, Code | 133,096,832 | 8 Nvidia L20 | Oct 28, 2024 |
| 2 | MPLP | No | 0.9072 ± 0.0012 | 0.9074 ± 0.0011 | Kaiwen Dong (University of Notre Dame) | Paper, Code | 749,757,283 | NVIDIA A100 GPU (80G RAM) | Jan 16, 2024 |
| 3 | GraphGPT(SMTP) | No | 0.9055 ± 0.0016 | 0.9042 ± 0.0014 | Qifang Zhao (Alibaba) | Paper, Code | 46,784,128 | 8 Nvidia L20 | Oct 10, 2024 |
| 4 | SIEG | No | 0.9018 ± 0.0015 | 0.9011 ± 0.0015 | anonymous20221001 (anonymous) | Paper, Code | 2,428,365 | P100 (16GB) | Oct 23, 2023 |
| 5 | CFG | No | 0.9003 ± 0.0007 | 0.8992 ± 0.0007 | Jingsong Lv (Zhuque@ZhejiangLab) | Paper, Code | 686,253 | NVIDIA A100 (80GB GPU) | Sep 30, 2023 |
| 6 | CFG | No | 0.8997 ± 0.0015 | 0.8987 ± 0.0011 | Jingsong LV (Zhuque@ZhejiangLab) | Paper, Code | 686,253 | NVIDIA A100 (80GB GPU) | Sep 14, 2023 |
| 7 | SIEG | No | 0.8987 ± 0.0018 | 0.8978 ± 0.0018 | anonymous20221001 (anonymous) | Paper, Code | 667,821 | P100 (16GB) | Jan 29, 2023 |
| 8 | SIEG | No | 0.8957 ± 0.0010 | 0.8948 ± 0.0008 | anonymous20221001 (anonymous) | Paper, Code | 256,802 | P100 | Dec 25, 2022 |
| 9 | GCN + Heuristic Encoding | No | 0.8891 ± 0.0005 | 0.8892 ± 0.0005 | Shuming Liang (UTS) | Paper, Code | 372,674 | NVIDIA A100 (40G) | Nov 24, 2024 |
| 9 | NGNN + SEAL | No | 0.8891 ± 0.0022 | 0.8879 ± 0.0022 | Yakun Song(DGL) | Paper, Code | 1,134,402 | Tesla T4(16GB) | Nov 6, 2022 |
| 10 | SUREL | No | 0.8883 ± 0.0018 | 0.8891 ± 0.0021 | Haoteng Yin (Purdue University) | Paper, Code | 79,617 | Quadro RTX 6000 | Mar 18, 2022 |
| 11 | S3GRL (PoS Plus) | No | 0.8814 ± 0.0008 | 0.8809 ± 0.0074 | Paul Louis (Ontario Tech University) | Paper, Code | 142,275,001 | GeForce GTX 1080 Ti (11 GB) | May 2, 2023 |
| 12 | BUDDY | No | 0.8796 ± 0.0008 | 0.8793 ± 0.0008 | Ben Chamberlain | Paper, Code | 166,531 | Tesla K80 (12GB GPU) | Aug 29, 2023 |
| 13 | SEAL | No | 0.8767 ± 0.0032 | 0.8757 ± 0.0031 | Muhan Zhang | Paper, Code | 260,802 | Tesla V100 (32GB GPU) | Feb 12, 2021 |
| 14 | AGDN w/GraphSAINT | No | 0.8549 ± 0.0029 | 0.8556 ± 0.0033 | Chuxiong Sun | Paper, Code | 306,716 | Tesla V100 (16GB GPU) | Sep 2, 2022 |
| 15 | PLNLP | No | 0.8492 ± 0.0029 | 0.8490 ± 0.0031 | Zhitao Wang (WeChat@Tencent) | Paper, Code | 146,514,551 | Tesla-V100 (32G GPU) | Dec 7, 2021 |
| 16 | Full-batch GCN | No | 0.8474 ± 0.0021 | 0.8479 ± 0.0023 | Matthias Fey – OGB team | Paper, Code | 296,449 | Quadro RTX 8000 (48GB GPU) | Jan 4, 2021 |
| 17 | HPE - Pre-trained Initialized | No | 0.8432 ± 0.0003 | 0.8422 ± 0.0002 | Chi-Hung Chang (CFDA & CLIP Labs) | Paper, Code | 749,558,528 | CPU | Dec 7, 2021 |
| 18 | Full-batch GraphSAGE | No | 0.8260 ± 0.0036 | 0.8263 ± 0.0033 | Matthias Fey – OGB team | Paper, Code | 460,289 | Quadro RTX 8000 (48GB GPU) | Jan 4, 2021 |
| 19 | NeighborSampling (SAGE aggr) | No | 0.8044 ± 0.0010 | 0.8054 ± 0.0009 | Matthias Fey – OGB team | Paper, Code | 460,289 | GeForce RTX 2080 (11GB GPU) | Jan 4, 2021 |
| 20 | ClusterGCN (GCN aggr) | No | 0.8004 ± 0.0025 | 0.7994 ± 0.0025 | Matthias Fey – OGB team | Paper, Code | 296,449 | GeForce RTX 2080 (11GB GPU) | Jan 4, 2021 |
| 21 | GraphSAINT (GCN aggr) | No | 0.7985 ± 0.0040 | 0.7975 ± 0.0039 | Matthias Fey – OGB team | Paper, Code | 296,449 | GeForce RTX 2080 (11GB GPU) | Jan 4, 2021 |
| 22 | Node2vec | No | 0.6141 ± 0.0011 | 0.6124 ± 0.0011 | Matthias Fey – OGB team | Paper, Code | 374,911,105 | GeForce RTX 2080 (11GB GPU) | Jan 4, 2021 |
| 23 | Adamic Adar | No | 0.5189 ± 0.0000 | 0.5167 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
| 24 | Matrix Factorization | No | 0.5186 ± 0.0443 | 0.5181 ± 0.0436 | Matthias Fey – OGB team | Paper, Code | 281,113,505 | GeForce RTX 2080 (11GB GPU) | Jan 4, 2021 |
| 25 | Common Neighbor | No | 0.5147 ± 0.0000 | 0.5119 ± 0.0000 | Muhan Zhang | Paper, Code | 0 | GeForce RTX 2080S (8GB GPU) | Feb 12, 2021 |
Leaderboard for ogbl-wikikg2
The MRR score on the test and validation sets. The higher, the better.
Package: >=1.2.4
Deprecated ogbl-wikikg leaderboard can be found here.
| Rank | Method | Ext. data | Test MRR | Validation MRR | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | RelEns | No | 0.7392 ± 0.0011 | 0.7509 ± 0.0009 | Quanming Yao (EE Tsinghua) | Paper, Code | 2,176,767,622 | Tesla A100(80GB) | Oct 15, 2023 |
| 2 | StarGraph + TripleRE + Text | Yes | 0.7305 ± 0.0010 | 0.7442 ± 0.0006 | Liang Yao (Tencent) | Paper, Code | 1,927,395,330 | Tesla V100 (32GB) | Jan 4, 2023 |
| 3 | InterHT+ | No | 0.7293 ± 0.0018 | 0.7391 ± 0.0023 | Baoxin Wang (HFL) | Paper, Code | 156,332,770 | Tesla A100 (80GB) | Dec 23, 2022 |
| 4 | StarGraph + TripleRE | No | 0.7286 ± 0.0009 | 0.7335 ± 0.0011 | Linhui Feng (360AI) | Paper, Code | 93,039,522 | Tesla A100 (40GB) | Jan 3, 2023 |
| 5 | InterHT+ (256dim) | No | 0.7257 ± 0.0018 | 0.7370 ± 0.0022 | Baoxin Wang (HFL) | Paper, Code | 148,000,738 | Tesla A100 (40GB) | Dec 26, 2022 |
| 6 | StarGraph + TripleRE | No | 0.7201 ± 0.0011 | 0.7288 ± 0.0008 | Hongzhu Li (360AI) | Paper, Code | 86,762,146 | Tesla A100(40GB) | May 30, 2022 |
| 7 | CompoundE3D | No | 0.7006 ± 0.0011 | 0.7175 ± 0.0015 | Xiou Ge (USC) | Paper, Code | 750,662,700 | Tesla A100 (40GB GPU) | Dec 29, 2023 |
| 8 | TranS | No | 0.6939 ± 0.0011 | 0.7058 ± 0.0018 | Xuanyu Zhang (DXM AI) | Paper, Code | 38,430,804 | Tesla V100 (16GB) | Apr 19, 2022 |
| 9 | TranS | No | 0.6882 ± 0.0019 | 0.6988 ± 0.0006 | Xuanyu Zhang (DXM AI) | Paper, Code | 19,215,402 | Tesla V100 (16GB) | Apr 28, 2022 |
| 10 | TripleRE + NodePiece | No | 0.6866 ± 0.0014 | 0.6955 ± 0.0008 | Long Yu (360AI) | Paper, Code | 36,421,802 | Tesla A100(40GB) | Feb 24, 2022 |
| 11 | InterHT | No | 0.6779 ± 0.0018 | 0.6893 ± 0.0015 | Baoxin Wang (HFL) | Paper, Code | 19,215,402 | Tesla V100 (32GB) | Feb 10, 2022 |
| 12 | A*Net | No | 0.6711 ± 0.0045 | 0.6796 ± 0.0070 | Zhaocheng Zhu (Mila) | Paper, Code | 6,831,201 | Tesla A100 (40GB) | May 21, 2023 |
| 13 | TripleRE + NodePiece | No | 0.6582 ± 0.0020 | 0.6616 ± 0.0018 | Long Yu (360AI) | Paper, Code | 7,289,002 | Tesla A100(40GB) | Dec 25, 2021 |
| 14 | ComplEx-RP (50dim) | No | 0.6392 ± 0.0045 | 0.6561 ± 0.0070 | Yihong Chen (UCL NLP & FAIR London) | Paper, Code | 250,167,400 | Tesla V100 (32GB) | Nov 23, 2021 |
| 15 | TransERR | No | 0.6359 ± 0.0020 | 0.6518 ± 0.0012 | Jiang Li (IMU-AI) | Paper, Code | 500,441,802 | Tesla V100 (32GB) | Jun 26, 2023 |
| 16 | tripleRE | No | 0.5794 ± 0.0020 | 0.6045 ± 0.0024 | Long Yu (360AI) | Paper, Code | 500,763,337 | Tesla P40(24GB) | Dec 17, 2021 |
| 17 | NodePiece + AutoSF | No | 0.5703 ± 0.0035 | 0.5806 ± 0.0047 | Mikhail Galkin (Mila) | Paper, Code | 6,860,602 | Tesla V100 (32 GB) | Jul 17, 2021 |
| 18 | Rot-Pro | No | 0.5602 ± 0.0016 | 0.5740 ± 0.0008 | Tengwei Song(BUAA-NLSDE) | Paper, Code | 1,000,669,602 | Tesla V100(32GB) | Sep 5, 2022 |
| 19 | AutoSF | No | 0.5458 ± 0.0052 | 0.5510 ± 0.0063 | Yongqi Zhang (4Paradigm) | Paper, Code | 500,227,800 | Quadro RTX 8000 (45GB GPU) | Apr 2, 2021 |
| 20 | PairRE (200dim) | No | 0.5208 ± 0.0027 | 0.5423 ± 0.0020 | Linlin Chao | Paper, Code | 500,334,800 | Tesla P100 (16GB GPU) | Jan 28, 2021 |
| 21 | RotatE (250dim) | No | 0.4332 ± 0.0025 | 0.4353 ± 0.0028 | Hongyu Ren – OGB team | Paper, Code | 1,250,435,750 | Quadro RTX 8000 (45GB GPU) | Jan 23, 2021 |
| 22 | TransE (500dim) | No | 0.4256 ± 0.0030 | 0.4272 ± 0.0030 | Hongyu Ren – OGB team | Paper, Code | 1,250,569,500 | Quadro RTX 8000 (45GB GPU) | Jan 23, 2021 |
| 23 | ComplEx (250dim) | No | 0.4027 ± 0.0027 | 0.3759 ± 0.0016 | Hongyu Ren – OGB team | Paper, Code | 1,250,569,500 | Quadro RTX 8000 (45GB GPU) | Jan 23, 2021 |
| 24 | ComplEx (50dim) | No | 0.3804 ± 0.0022 | 0.3534 ± 0.0052 | Hongyu Ren – OGB team | Paper, Code | 250,113,900 | GeForce RTX 2080 (11GB GPU) | Jan 23, 2021 |
| 25 | DistMult (500dim) | No | 0.3729 ± 0.0045 | 0.3506 ± 0.0042 | Hongyu Ren – OGB team | Paper, Code | 1,250,569,500 | Quadro RTX 8000 (45GB GPU) | Jan 23, 2021 |
| 26 | DistMult (100dim) | No | 0.3447 ± 0.0082 | 0.3150 ± 0.0088 | Hongyu Ren – OGB team | Paper, Code | 250,113,900 | GeForce RTX 2080 (11GB GPU) | Jan 23, 2021 |
| 27 | TransE (100dim) | No | 0.2622 ± 0.0045 | 0.2465 ± 0.0020 | Hongyu Ren – OGB team | Paper, Code | 250,113,900 | GeForce RTX 2080 (11GB GPU) | Jan 23, 2021 |
| 28 | RotatE (50dim) | No | 0.2530 ± 0.0034 | 0.2250 ± 0.0035 | Hongyu Ren – OGB team | Paper, Code | 250,087,150 | GeForce RTX 2080 (11GB GPU) | Jan 23, 2021 |
Leaderboard for ogbl-biokg
The MRR score on the test and validation sets. The higher, the better.
Package: >=1.2.0
| Rank | Method | Ext. data | Test MRR | Validation MRR | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | RelEns | No | 0.9618 ± 0.0002 | 0.9627 ± 0.0004 | Quanming Yao (EE Tsinghua) | Paper, Code | 849,427,106 | Tesla A100 (80GB) | Oct 15, 2023 |
| 2 | ComplEx^2 | No | 0.8583 ± 0.0005 | 0.8592 ± 0.0004 | Lorenzo Loconte (Univ. of Edinburgh) | Paper, Code | 187,648,000 | NVIDIA RTX A6000 (48GB) | Oct 28, 2023 |
| 3 | UniBi | No | 0.8550 ± 0.0003 | 0.8553 ± 0.0001 | Ruilin Luo (THU) | Paper, Code | 181,654,170 | RTX 4090 (24GB) | Dec 2, 2023 |
| 4 | AutoBLM-KGBench | No | 0.8536 ± 0.0003 | 0.8548 ± 0.0002 | Yongqi Zhang (4Paradigm && EE Tsinghua) | Paper, Code | 192,047,104 | Tesla A100 (80G) | Jun 17, 2022 |
| 5 | ComplEx-RP (1000dim) | No | 0.8492 ± 0.0002 | 0.8497 ± 0.0002 | Yihong Chen (UCL NLP & FAIR London) | Paper, Code | 187,750,000 | Tesla V100 (16GB) | Nov 23, 2021 |
| 6 | TripleRE | No | 0.8348 ± 0.0007 | 0.8360 ± 0.0006 | LongYu(360AI) | Paper, Code | 469,630,002 | Tesla A100(40GB) | Jan 28, 2022 |
| 7 | AutoSF | No | 0.8309 ± 0.0008 | 0.8317 ± 0.0007 | Yongqi Zhang (4Paradigm) | Paper, Code | 93,824,000 | GeForce RTX 2080 (11GB GPU) | Apr 2, 2021 |
| 8 | PairRE | No | 0.8164 ± 0.0005 | 0.8172 ± 0.0005 | LinlinChao (AntGroup KG&NLP) | Paper, Code | 187,750,000 | Tesla P100 (16GB GPU) | Nov 9, 2020 |
| 9 | ComplEx | No | 0.8095 ± 0.0007 | 0.8105 ± 0.0001 | Hongyu Ren – OGB team | Paper, Code | 187,648,000 | GeForce RTX 2080 (11GB GPU) | Jun 10, 2020 |
| 10 | DistMult | No | 0.8043 ± 0.0003 | 0.8055 ± 0.0003 | Hongyu Ren – OGB team | Paper, Code | 187,648,000 | GeForce RTX 2080 (11GB GPU) | Jun 10, 2020 |
| 11 | RotatE | No | 0.7989 ± 0.0004 | 0.7997 ± 0.0002 | Hongyu Ren – OGB team | Paper, Code | 187,597,000 | GeForce RTX 2080 (11GB GPU) | Jun 10, 2020 |
| 12 | TransE | No | 0.7452 ± 0.0004 | 0.7456 ± 0.0003 | Hongyu Ren – OGB team | Paper, Code | 187,648,000 | GeForce RTX 2080 (11GB GPU) | Jun 10, 2020 |
Leaderboard for ogbl-vessel
The ROC-AUC score on the test and validation sets. The higher, the better.
Package: >=1.3.4
| Rank | Method | Ext. data | Test ROC-AUC | Validation ROC-AUC | Contact | References | #Params | Hardware | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | GAV | No | 0.9838 ± 0.0002 | 0.9837 ± 0.0002 | Bastian Wittmann (UZH) | Paper, Code | 8,194 | Tesla V100 (32 GB GPU) | Mar 28, 2023 |
| 2 | SIEG(SPG) | No | 0.9687 ± 0.0001 | 0.9687 ± 0.0001 | anonymous20221001 (anonymous) | Paper, Code | 1,306,956 | P100 (16GB) | Aug 15, 2023 |
| 3 | MLP+DiffPredictor | No | 0.9414 ± 0.0001 | 0.9415 ± 0.0002 | Chuxiong Sun (CTRI) | Paper, Code | 16,961 | Tesla V100 (16GB GPU) | Aug 7, 2023 |
| 4 | SIEG w/o NF | No | 0.8798 ± 0.0100 | 0.8803 ± 0.0100 | anonymous20221001 (anonymous) | Paper, Code | 752,716 | P100 (16GB) | Mar 27, 2023 |
| 5 | SUREL+ | No | 0.8496 ± 0.0068 | 0.8489 ± 0.0067 | Haoteng Yin (Purdue University) | Paper, Code | 56,353 | Quadro RTX 6000 (24GB) | Feb 28, 2023 |
| 6 | SIEG | No | 0.8307 ± 0.0044 | 0.8315 ± 0.0044 | anonymous20221001 (anonymous) | Paper, Code | 407,338 | P100 (16GB) | Jan 29, 2023 |
| 7 | SIEG | No | 0.8249 ± 0.0041 | 0.8255 ± 0.0040 | anonymous20221001 (anonymous) | Paper, Code | 594,498 | P100 (16GB) | Oct 1, 2022 |
| 8 | SEAL (no-xfeat) | No | 0.8077 ± 0.0001 | 0.8073 ± 0.0001 | Huixuan Chi (Meituan) | Paper, Code | 43,714 | Tesla V100 (32GB) | Aug 28, 2022 |
| 9 | S3GRL (PoS Plus) | Yes | 0.8056 ± 0.0006 | 0.8065 ± 0.0022 | Paul Louis (Ontario Tech University) | Paper, Code | 2,382,849 | GeForce GTX 1080 Ti (11 GB) | May 2, 2023 |
| 10 | SEAL | No | 0.8050 ± 0.0021 | 0.8053 ± 0.0022 | Julian McGinnis (TUM) | Paper, Code | 172,610 | Quadro RTX 8000 Ti (48GB GPU) | Aug 24, 2022 |
| 11 | LRGA | No | 0.5415 ± 0.0437 | 0.5418 ± 0.0439 | Yuanchen Bei (Meituan) | Paper, Code | 26,577 | Tesla A100 (80GB) | Aug 31, 2022 |
| 12 | SGC | No | 0.5009 ± 0.0011 | 0.5010 ± 0.0011 | Hao Xu(Meituan) | Paper, Code | 897 | Tesla V100(32GB) | Aug 24, 2022 |
| 13 | SAGE+JKNet | No | 0.5001 ± 0.0007 | 0.5014 ± 0.0004 | Huixuan Chi (Meituan) | Paper, Code | 273 | Tesla V100 (32GB) | Aug 22, 2022 |
| 14 | Matrix Factorization | No | 0.4997 ± 0.0005 | 0.4999 ± 0.0006 | Julian McGinnis (TUM) | Paper, Code | 8,641 | Quadro RTX 8000Ti (48GB GPU) | Aug 18, 2022 |
| 15 | GraphSAGE | No | 0.4989 ± 0.0678 | 0.4993 ± 0.0676 | Julian McGinnis (TUM) | Paper, Code | 396,289 | Quadro RTX 8000 Ti (48GB GPU) | Aug 20, 2022 |
| 16 | GCN w/Node2Vec | No | 0.4954 ± 0.0057 | 0.4960 ± 0.0061 | Julian McGinnis (TUM) | Paper, Code | 226,744,513 | Quadro RTX 8000 Ti (48GB GPU) | Aug 21, 2022 |
| 17 | Common Neighbors | No | 0.4849 ± 0.0000 | 0.4850 ± 0.0000 | Julian McGinnis (TUM) | Paper, Code | 0 | Quadro RTX 8000 Ti (48GB GPU) | Aug 20, 2022 |
| 17 | Adamic Adar | No | 0.4849 ± 0.0000 | 0.4849 ± 0.0000 | Julian McGinnis (TUM) | Paper, Code | 0 | Quadro RTX 8000 Ti (48GB GPU) | Aug 20, 2022 |
| 18 | MLP | No | 0.4794 ± 0.0133 | 0.4801 ± 0.0132 | Julian McGinnis (TUM) | Paper, Code | 1,037,577 | Quadro RTX 8000 Ti (48GB GPU) | Aug 19, 2022 |
| 19 | GraphSAGE w/Node2VEc | No | 0.4735 ± 0.0136 | 0.4736 ± 0.0136 | Julian McGinnis (TUM) | Paper, Code | 226,892,737 | Quadro RTX 8000 Ti (48GB GPU) | Aug 21, 2022 |
| 20 | GCN | No | 0.4353 ± 0.0961 | 0.4349 ± 0.0961 | Julian McGinnis (TUM) | Paper, Code | 396,289 | Quadro RTX 8000 Ti (48GB GPU) | Aug 20, 2022 |