Rethinking Multi-Label Node Classification: Do Tuned Classic GNNs Suffice?

Shengzhong Zhang; Yuxuan Xiao

arxiv: 2605.01403 · v1 · submitted 2026-05-02 · 💻 cs.LG

Rethinking Multi-Label Node Classification: Do Tuned Classic GNNs Suffice?

Yuxuan Xiao , Shengzhong Zhang This is my paper

Pith reviewed 2026-05-09 15:19 UTC · model grok-4.3

classification 💻 cs.LG

keywords multi-label node classificationgraph neural networksGNN baselineshyperparameter tuningnode classificationmulti-label learninggraph learning

0 comments

The pith

Tuned classic GNN backbones outperform specialized label-aware methods for multi-label node classification on four of five benchmarks.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper asks whether the advantages of complex label-aware designs in multi-label node classification truly come from modeling node-label interactions or from weak baselines. It tests this by applying standard optimization techniques such as normalization, dropout, and residual connections to classic full-graph GNNs including GCN, SSGConv, and GCNII. Experiments on five representative benchmark datasets show the tuned models exceed representative specialized methods on four datasets and reach state-of-the-art performance in multiple settings. This indicates that hyperparameter tuning and basic architectural refinements are more decisive than previously emphasized. The work urges future multi-label graph learning studies to adopt stronger baseline comparisons.

Core claim

The authors establish that carefully tuned classic full-graph GNNs already serve as strong solutions for multi-label node classification, outperforming representative specialized methods on four out of five benchmark datasets and achieving state-of-the-art results in multiple settings.

What carries the argument

Systematic hyperparameter optimization of classic GNN backbones with normalization layers, dropout regularization, and residual connections.

If this is right

Tuned classic GNNs surpass specialized methods on four of the five benchmark datasets.
State-of-the-art performance is reached in multiple experimental settings using only optimized standard backbones.
Tuning effort and simple techniques explain much of the reported gains from complex designs.
Future research on multi-label graph learning requires more rigorous strong-baseline evaluation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Label-interaction modeling may contribute limited additional value once baselines receive comparable tuning.
Similar baseline re-evaluations could apply to other graph tasks where new architectures are introduced without exhaustive tuning.
Practitioners may obtain competitive results with simpler models by prioritizing optimization over novel designs.

Load-bearing premise

The specialized label-aware methods were not given equivalent hyperparameter optimization effort and that the chosen normalization, dropout, and residual techniques confer no unfair advantage to the classic backbones.

What would settle it

Re-optimizing the specialized label-aware methods with the same hyperparameter search budget and architectural tweaks, then checking whether they match or exceed the tuned classic GNN performance on the same five datasets.

read the original abstract

Multi-label node classification (MLNC) has recently been addressed by increasingly complex label-aware designs that explicitly model node-label interactions and inter-label dependencies.However, it remains unclear whether the advantages of these methods truly stem from their specialized designs, or simply from insufficiently optimized baselines. In this paper, we revisit MLNC from a strong-baseline perspective and investigate whether carefully tuned classic full-graph GNNs can already serve as strong solutions to this task. We systematically study several representative backbones, including GCN, SSGConv, and GCNII, and optimize them using standard yet effective techniques such as normalization, dropout, and residual connections. Experiments on five representative benchmark datasets show that our tuned baselines outperform representative specialized methods on four datasets and achieve state-of-the-art performance in multiple settings. These results indicate that careful tuning of classic backbones is a highly influential but often overlooked factor in MLNC, and highlight the need for more rigorous strong-baseline evaluation in future research on multi-label graph learning.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Tuned classic GNNs beat specialized MLNC methods on four of five benchmarks, but the comparison fairness on tuning effort is the main open question.

read the letter

Tuned versions of GCN, SSGConv, and GCNII already outperform representative specialized label-aware methods on four of the five standard benchmarks. That is the core observation worth noting right away. The paper adds normalization, dropout, and residual connections to these backbones and reports the resulting numbers directly against published results for the more complex designs. This produces state-of-the-art marks in several settings and shows that careful use of ordinary techniques can close or reverse the apparent gap created by newer architectures. The experiments cover five representative datasets, which gives the comparison a reasonable scope. The work is useful because it supplies concrete baseline numbers that future papers in multi-label node classification can reference when they claim gains from label-aware modeling. It also serves as a reminder that optimization details often matter more than architectural novelty in this area. The soft spot is the treatment of the specialized methods. The paper applies the standard tricks to the classics but does not indicate that the same hyperparameter search budget or augmentations were given to the competitors. If those methods were evaluated only in their original reported configurations, the performance difference could narrow once both sides receive equivalent effort. That does not make the classic-GNN results wrong, but it limits how strongly one can conclude that specialized designs add little value. This paper is for graph ML researchers who run multi-label node classification experiments and want to know what a properly tuned baseline actually achieves. A reader focused on evaluation practices will find the numbers helpful. It deserves a serious referee because the question about baseline strength is substantive and the empirical results are specific, even though clearer documentation of the tuning process for all methods would make the case tighter. I would send it to peer review rather than desk reject.

Referee Report

2 major / 2 minor

Summary. The paper claims that carefully tuned classic full-graph GNN backbones (GCN, SSGConv, GCNII) augmented with normalization, dropout, and residual connections outperform representative specialized label-aware methods for multi-label node classification on four of five benchmark datasets and achieve state-of-the-art results in multiple settings. It concludes that insufficient baseline optimization, rather than inherent limitations of classic designs, may explain the reported advantages of complex label-aware architectures.

Significance. If the comparisons prove fair and reproducible, this would be a significant contribution by demonstrating that standard optimization techniques can close or reverse performance gaps in MLNC without specialized label modeling. The multi-dataset empirical evaluation provides a useful reference point for the field and could encourage stronger baseline practices, reducing unnecessary architectural complexity. The emphasis on full-graph GNNs with explicit augmentations is a concrete strength that, if detailed, supports falsifiable claims about tuning impact.

major comments (2)

[Experiments / Results] Experimental results (as summarized in the abstract and presumably detailed in the results section/tables): the central claim of outperformance on four datasets rests on comparing the authors' tuned classics against published numbers for specialized methods, but no information is given on whether equivalent hyperparameter search budgets, ranges, or the same normalization/dropout/residual augmentations were applied to those specialized baselines. This directly affects whether the gap reflects design differences or optimization disparity.
[Methodology] Methodology section: while the paper systematically optimizes the classic backbones with normalization, dropout, and residuals, there is no corresponding description, re-implementation details, or ablation confirming that the representative specialized label-aware methods received comparable treatment. Without this, the inference that specialized designs add little value beyond tuning cannot be fully supported.

minor comments (2)

[Abstract] The abstract states 'state-of-the-art performance in multiple settings' without naming the specific datasets or metrics where this occurs; adding this would improve precision and allow readers to quickly locate the supporting tables.
[Experiments] No mention of statistical significance testing or variance across runs is referenced in the provided summary of results; including error bars or p-values would strengthen the outperformance claims.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback, which highlights important aspects of our experimental design and helps strengthen the manuscript. We address each major comment below and describe the revisions we will make.

read point-by-point responses

Referee: [Experiments / Results] Experimental results (as summarized in the abstract and presumably detailed in the results section/tables): the central claim of outperformance on four datasets rests on comparing the authors' tuned classics against published numbers for specialized methods, but no information is given on whether equivalent hyperparameter search budgets, ranges, or the same normalization/dropout/residual augmentations were applied to those specialized baselines. This directly affects whether the gap reflects design differences or optimization disparity.

Authors: We appreciate the referee's point on the nature of the comparison. The results for specialized label-aware methods are taken directly from the numbers reported in their original papers, which reflect the optimization efforts described therein. Our work demonstrates that classic full-graph GNNs, when systematically tuned with normalization, dropout, and residual connections, can match or exceed these published figures on four of five datasets. This supports the claim that baseline optimization is a key overlooked factor. We agree a fully controlled re-implementation and re-tuning of specialized methods would provide an even stronger comparison. We will revise the experimental setup section to explicitly state that specialized results are as-published and add a discussion paragraph on the value of standardized tuning protocols for future MLNC research. revision: partial
Referee: [Methodology] Methodology section: while the paper systematically optimizes the classic backbones with normalization, dropout, and residuals, there is no corresponding description, re-implementation details, or ablation confirming that the representative specialized label-aware methods received comparable treatment. Without this, the inference that specialized designs add little value beyond tuning cannot be fully supported.

Authors: We agree that the methodology requires greater transparency on this point. Our tuning procedure for the classic GNN backbones (GCN, SSGConv, GCNII) is detailed, including hyperparameter ranges and the application of augmentations, but we did not re-implement or apply equivalent tuning to the specialized methods. We will expand the methodology section with a dedicated paragraph clarifying the sources of all comparison results and noting that no additional optimization was performed on the specialized approaches beyond their original reports. We will also add a brief ablation on the contribution of the augmentations to our tuned classics to better isolate their effect. revision: yes

Circularity Check

0 steps flagged

No circularity in empirical baseline comparison

full rationale

The paper conducts an empirical evaluation of tuned classic GNN backbones (GCN, SSGConv, GCNII) against specialized multi-label node classification methods across five benchmark datasets. No mathematical derivation, predictive chain, or self-referential definition is present; results are reported directly from hyperparameter-tuned experiments using standard techniques such as normalization, dropout, and residuals. The central claim rests on observed performance numbers rather than any reduction of outputs to inputs by construction, self-citation of uniqueness theorems, or smuggling of ansatzes. This is a standard empirical comparison paper whose validity hinges on experimental fairness, not internal logical circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is an empirical benchmarking study with no mathematical derivations, new theoretical constructs, or postulated entities; all claims rest on experimental comparisons of existing models.

pith-pipeline@v0.9.0 · 5473 in / 1075 out tokens · 35350 ms · 2026-05-09T15:19:30.251942+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

67 extracted references · 67 canonical work pages

[1]

Langley , title =

P. Langley , title =. Proceedings of the 17th International Conference on Machine Learning (ICML 2000) , address =. 2000 , pages =

work page 2000
[2]

T. M. Mitchell. The Need for Biases in Learning Generalizations. 1980

work page 1980
[3]

M. J. Kearns , title =

work page
[4]

Machine Learning: An Artificial Intelligence Approach, Vol. I. 1983

work page 1983
[5]

R. O. Duda and P. E. Hart and D. G. Stork. Pattern Classification. 2000

work page 2000
[6]

Suppressed for Anonymity , author=

work page
[7]

Newell and P

A. Newell and P. S. Rosenbloom. Mechanisms of Skill Acquisition and the Law of Practice. Cognitive Skills and Their Acquisition. 1981

work page 1981
[8]

A. L. Samuel. Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development. 1959

work page 1959
[9]

2018 , eprint=

Learning to Navigate in Cities Without a Map , author=. 2018 , eprint=

work page 2018
[10]

Vlahavas , editor =

Grigorios Tsoumakas and Ioannis Katakis and Ioannis P. Vlahavas , editor =. Mining Multi-label Data , booktitle =. 2010 , url =

work page 2010
[11]

Proceedings of the ACM Symposium on Document Engineering 2023, DocEng 2023, Limerick, Ireland, August 22-25, 2023 , pages =

Haytame Fallah and Emmanuel Bruno and Patrice Bellot and Elisabeth Murisasco , title =. Proceedings of the ACM Symposium on Document Engineering 2023, DocEng 2023, Limerick, Ireland, August 22-25, 2023 , pages =. 2023 , url =

work page 2023
[12]

Scientometrics , volume =

Arousha Haghighian Roudsari and Jafar Afshar and Wookey Lee and Suan Lee , title =. Scientometrics , volume =. 2022 , url =

work page 2022
[13]

Priyadharshini and A

M. Priyadharshini and A. Faritha Banu and Bhisham Sharma and Subrata Chowdhury and Khaled M. Rabie and Thokozani Shongwe , title =. Sensors , volume =. 2023 , url =

work page 2023
[14]

Multilabel classification of medical concepts for patient clinical profile identification , journal =

Christel G. Multilabel classification of medical concepts for patient clinical profile identification , journal =. 2022 , url =

work page 2022
[15]

A new Classifier Chain method of BERT Models For Multi-label Classification of Arabic Abusive Language on Social Media , booktitle =

Salma Abid Azzi and Chiraz Ben Othmane Zribi , editor =. A new Classifier Chain method of BERT Models For Multi-label Classification of Arabic Abusive Language on Social Media , booktitle =. 2023 , url =

work page 2023
[16]

Pattern Recognit

Min-Ling Zhang and Zhi-Hua Zhou , title =. Pattern Recognit. , volume =. 2007 , url =

work page 2007
[17]

Vlahavas , editor =

Grigorios Tsoumakas and Ioannis P. Vlahavas , editor =. Random k-Labelsets: An Ensemble Method for Multilabel Classification , booktitle =. 2007 , url =

work page 2007
[18]

Multi-label Ranking with LSTM² for Document Classification , booktitle =

Yan Yan and Xu-Cheng Yin and Chun Yang and Bo-Wen Zhang and Hong-Wei Hao , editor =. Multi-label Ranking with LSTM² for Document Classification , booktitle =. 2016 , url =

work page 2016
[19]

Transactions on Machine Learning Research , issn=

Multi-label Node Classification On Graph-Structured Data , author=. Transactions on Machine Learning Research , issn=. 2023 , url=

work page 2023
[20]

Open Graph Benchmark: Datasets for Machine Learning on Graphs , booktitle =

Weihua Hu and Matthias Fey and Marinka Zitnik and Yuxiao Dong and Hongyu Ren and Bowen Liu and Michele Catasta and Jure Leskovec , editor =. Open Graph Benchmark: Datasets for Machine Learning on Graphs , booktitle =. 2020 , url =

work page 2020
[21]

Sheng , title =

Cangqi Zhou and Hui Chen and Jing Zhang and Qianmu Li and Dianming Hu and Victor S. Sheng , title =. Expert Syst. Appl. , volume =. 2021 , url =

work page 2021
[22]

Kipf and Max Welling , title =

Thomas N. Kipf and Max Welling , title =. CoRR , volume =. 2016 , url =

work page 2016
[23]

Graph Attention Networks , booktitle =

Petar Velickovic and Guillem Cucurull and Arantxa Casanova and Adriana Romero and Pietro Li. Graph Attention Networks , booktitle =. 2018 , url =

work page 2018
[24]

node2vec: Scalable Feature Learning for Networks , booktitle =

Aditya Grover and Jure Leskovec , editor =. node2vec: Scalable Feature Learning for Networks , booktitle =. 2016 , url =

work page 2016
[25]

BMC Bioinform

Guangyu Cui and Chao Fang and Kyungsook Han , title =. BMC Bioinform. , volume =. 2012 , url =

work page 2012
[26]

Correlation-Aware Graph Convolutional Networks for Multi-Label Node Classification , booktitle =

Yuanchen Bei and Weizhi Chen and Hao Chen and Sheng Zhou and Carl Ji Yang and Jiapei Fan and Longtao Huang and Jiajun Bu , editor =. Correlation-Aware Graph Convolutional Networks for Multi-Label Node Classification , booktitle =. 2025 , url =

work page 2025
[27]

Zhi-Peng Li and Siguo Wang and Qinhu Zhang and Yi-Jie Pan and Naian Xiao and Jiayang Guo and Chang-An Yuan and Wen-Jian Liu and De-Shuang Huang , title =. Artif. Intell. Rev. , volume =. 2025 , url =

work page 2025
[28]

Dhillon and Yuqiang Guan and Brian Kulis , title =

Inderjit S. Dhillon and Yuqiang Guan and Brian Kulis , title =. IEEE Trans. Pattern Anal. Mach. Intell. , volume =. 2007 , url =

work page 2007
[29]

CoRR , volume =

Mikael Henaff and Joan Bruna and Yann LeCun , title =. CoRR , volume =. 2015 , url =

work page 2015
[30]

Self-Attention Graph Pooling , booktitle =

Junhyun Lee and Inyeop Lee and Jaewoo Kang , editor =. Self-Attention Graph Pooling , booktitle =. 2019 , url =

work page 2019
[31]

Graph U-Nets , booktitle =

Hongyang Gao and Shuiwang Ji , editor =. Graph U-Nets , booktitle =. 2019 , url =

work page 2019
[32]

Structure-Feature based Graph Self-adaptive Pooling , booktitle =

Liang Zhang and Xudong Wang and Hongsheng Li and Guangming Zhu and Peiyi Shen and Ping Li and Xiaoyuan Lu and Syed Afaq Ali Shah and Mohammed Bennamoun , editor =. Structure-Feature based Graph Self-adaptive Pooling , booktitle =. 2020 , url =

work page 2020
[33]

Deep Convolutional Ranking for Multilabel Image Annotation , booktitle =

Yunchao Gong and Yangqing Jia and Thomas Leung and Alexander Toshev and Sergey Ioffe , editor =. Deep Convolutional Ranking for Multilabel Image Annotation , booktitle =. 2014 , url =

work page 2014
[34]

2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016 , pages =

Jiang Wang and Yi Yang and Junhua Mao and Zhiheng Huang and Chang Huang and Wei Xu , title =. 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016 , pages =. 2016 , url =

work page 2016
[35]

The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025 , publisher =

Yifei Sun and Zemin Liu and Bryan Hooi and Yang Yang and Rizal Fathony and Jia Chen and Bingsheng He , title =. The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025 , publisher =. 2025 , url =

work page 2025
[36]

Zheng and Kevin Chen-Chuan Chang , title =

Hongyun Cai and Vincent W. Zheng and Kevin Chen-Chuan Chang , title =. 2018 , url =

work page 2018
[37]

Shunxin Xiao and Shiping Wang and Yuanfei Dai and Wenzhong Guo , title =. Mach. Vis. Appl. , volume =. 2022 , url =

work page 2022
[38]

Attentive Recurrent Social Recommendation , booktitle =

Peijie Sun and Le Wu and Meng Wang , editor =. Attentive Recurrent Social Recommendation , booktitle =. 2018 , url =

work page 2018
[39]

Prasanna , title =

Hanqing Zeng and Hongkuan Zhou and Ajitesh Srivastava and Rajgopal Kannan and Viktor K. Prasanna , title =. 8th International Conference on Learning Representations,. 2020 , url =

work page 2020
[40]

2019 , url =

Thibaut Durand and Nazanin Mehrasa and Greg Mori , title =. 2019 , url =

work page 2019
[41]

Lin Xiao and Pengyu Xu and Liping Jing and Uchenna Akujuobi and Xiangliang Zhang , title =. Inf. Sci. , volume =. 2022 , url =

work page 2022
[42]

Semi-supervised Graph Embedding for Multi-label Graph Node Classification , booktitle =

Kaisheng Gao and Jing Zhang and Cangqi Zhou , editor =. Semi-supervised Graph Embedding for Multi-label Graph Node Classification , booktitle =. 2019 , url =

work page 2019
[43]

2022 , doi =

Ziwei Zhang and Peng Cui and Wenwu Zhu , title =. 2022 , doi =

work page 2022
[44]

Hamilton and Rex Ying and Jure Leskovec , title =

William L. Hamilton and Rex Ying and Jure Leskovec , title =. 2018 , eprint =

work page 2018
[45]

2019 , eprint =

Keyulu Xu and Weihua Hu and Jure Leskovec and Stefanie Jegelka , title =. 2019 , eprint =

work page 2019
[46]

Hamilton and Jure Leskovec , title =

Rex Ying and Jiaxuan You and Christopher Morris and Xiang Ren and William L. Hamilton and Jure Leskovec , title =. 2019 , eprint =

work page 2019
[47]

Taylor and Mohamed R

Boris Knyazev and Graham W. Taylor and Mohamed R. Amer , title =. Proceedings of the 33rd International Conference on Neural Information Processing Systems , publisher =. 2019 , articleno =

work page 2019
[48]

, title =

Zhang, Jiawei and Yu, Philip S. , title =. 2018 , issue_date =. doi:10.1145/3229329.3229333 , journal =

work page doi:10.1145/3229329.3229333 2018
[49]

Hierarchical graph learning for protein–protein interaction , volume =

Gao, Ziqi and Jiang, Chenran and Zhang, Jiawen and Jiang, Xiaosen and Li, Lanqing and Zhao, Peilin and Yang, Huanming and Huang, Yong and Li, Jia , year =. Hierarchical graph learning for protein–protein interaction , volume =. Nature Communications , doi =

work page
[50]

Semi-Supervised Learning Literature Survey , volume =

Zhu, Xiaojin , year =. Semi-Supervised Learning Literature Survey , volume =

work page
[51]

Wu, Zonghan and Pan, Shirui and Chen, Fengwen and Long, Guodong and Zhang, Chengqi and Yu, Philip S. , year=. A Comprehensive Survey on Graph Neural Networks , volume=. IEEE Transactions on Neural Networks and Learning Systems , publisher=. doi:10.1109/tnnls.2020.2978386 , number=

work page doi:10.1109/tnnls.2020.2978386 2020
[52]

2022 , eprint=

Benchmarking Graph Neural Networks , author=. 2022 , eprint=

work page 2022
[53]

2024 , eprint=

Classic GNNs are Strong Baselines: Reassessing GNNs for Node Classification , author=. 2024 , eprint=

work page 2024
[54]

2021 , eprint=

Bag of Tricks for Node Classification with Graph Neural Networks , author=. 2021 , eprint=

work page 2021
[55]

International Conference on Learning Representations , year=

Simple Spectral Graph Convolution , author=. International Conference on Learning Representations , year=

work page
[56]

2020 , eprint=

Simple and Deep Graph Convolutional Networks , author=. 2020 , eprint=

work page 2020
[57]

2016 , eprint=

Layer Normalization , author=. 2016 , eprint=

work page 2016
[58]

2015 , eprint=

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , author=. 2015 , eprint=

work page 2015
[59]

Srivastava, Nitish and Hinton, Geoffrey and Krizhevsky, Alex and Sutskever, Ilya and Salakhutdinov, Ruslan , title =. J. Mach. Learn. Res. , month = jan, pages =. 2014 , issue_date =

work page 2014
[60]

2015 , eprint=

Deep Residual Learning for Image Recognition , author=. 2015 , eprint=

work page 2015
[61]

InProceedings of the 30th ACM International Conference on Information & Knowledge Management (CIKM ’21)

Song, Zixing and Meng, Ziqiao and Zhang, Yifei and King, Irwin , title =. 2021 , isbn =. doi:10.1145/3459637.3482391 , booktitle =

work page doi:10.1145/3459637.3482391 2021
[62]

A Review on Multi-Label Learning Algorithms , year=

Zhang, Min-Ling and Zhou, Zhi-Hua , journal=. A Review on Multi-Label Learning Algorithms , year=

work page
[63]

2020 , eprint=

Graph-Revised Convolutional Network , author=. 2020 , eprint=

work page 2020
[64]

Zaki , editor =

Yu Chen and Lingfei Wu and Mohammed J. Zaki , editor =. Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings , booktitle =. 2020 , url =

work page 2020
[65]

Towards Unsupervised Deep Graph Structure Learning , booktitle =

Yixin Liu and Yu Zheng and Daokun Zhang and Hongxu Chen and Hao Peng and Shirui Pan , editor =. Towards Unsupervised Deep Graph Structure Learning , booktitle =. 2022 , url =. doi:10.1145/3485447.3512186 , timestamp =

work page doi:10.1145/3485447.3512186 2022
[66]

2022 , url =

Hongwei Wang and Jure Leskovec , title =. 2022 , url =. doi:10.1145/3490478 , timestamp =

work page doi:10.1145/3490478 2022
[67]

Kingma and Jimmy Ba , editor =

Diederik P. Kingma and Jimmy Ba , editor =. Adam:. 3rd International Conference on Learning Representations,. 2015 , url =

work page 2015

[1] [1]

Langley , title =

P. Langley , title =. Proceedings of the 17th International Conference on Machine Learning (ICML 2000) , address =. 2000 , pages =

work page 2000

[2] [2]

T. M. Mitchell. The Need for Biases in Learning Generalizations. 1980

work page 1980

[3] [3]

M. J. Kearns , title =

work page

[4] [4]

Machine Learning: An Artificial Intelligence Approach, Vol. I. 1983

work page 1983

[5] [5]

R. O. Duda and P. E. Hart and D. G. Stork. Pattern Classification. 2000

work page 2000

[6] [6]

Suppressed for Anonymity , author=

work page

[7] [7]

Newell and P

A. Newell and P. S. Rosenbloom. Mechanisms of Skill Acquisition and the Law of Practice. Cognitive Skills and Their Acquisition. 1981

work page 1981

[8] [8]

A. L. Samuel. Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development. 1959

work page 1959

[9] [9]

2018 , eprint=

Learning to Navigate in Cities Without a Map , author=. 2018 , eprint=

work page 2018

[10] [10]

Vlahavas , editor =

Grigorios Tsoumakas and Ioannis Katakis and Ioannis P. Vlahavas , editor =. Mining Multi-label Data , booktitle =. 2010 , url =

work page 2010

[11] [11]

Proceedings of the ACM Symposium on Document Engineering 2023, DocEng 2023, Limerick, Ireland, August 22-25, 2023 , pages =

Haytame Fallah and Emmanuel Bruno and Patrice Bellot and Elisabeth Murisasco , title =. Proceedings of the ACM Symposium on Document Engineering 2023, DocEng 2023, Limerick, Ireland, August 22-25, 2023 , pages =. 2023 , url =

work page 2023

[12] [12]

Scientometrics , volume =

Arousha Haghighian Roudsari and Jafar Afshar and Wookey Lee and Suan Lee , title =. Scientometrics , volume =. 2022 , url =

work page 2022

[13] [13]

Priyadharshini and A

M. Priyadharshini and A. Faritha Banu and Bhisham Sharma and Subrata Chowdhury and Khaled M. Rabie and Thokozani Shongwe , title =. Sensors , volume =. 2023 , url =

work page 2023

[14] [14]

Multilabel classification of medical concepts for patient clinical profile identification , journal =

Christel G. Multilabel classification of medical concepts for patient clinical profile identification , journal =. 2022 , url =

work page 2022

[15] [15]

A new Classifier Chain method of BERT Models For Multi-label Classification of Arabic Abusive Language on Social Media , booktitle =

Salma Abid Azzi and Chiraz Ben Othmane Zribi , editor =. A new Classifier Chain method of BERT Models For Multi-label Classification of Arabic Abusive Language on Social Media , booktitle =. 2023 , url =

work page 2023

[16] [16]

Pattern Recognit

Min-Ling Zhang and Zhi-Hua Zhou , title =. Pattern Recognit. , volume =. 2007 , url =

work page 2007

[17] [17]

Vlahavas , editor =

Grigorios Tsoumakas and Ioannis P. Vlahavas , editor =. Random k-Labelsets: An Ensemble Method for Multilabel Classification , booktitle =. 2007 , url =

work page 2007

[18] [18]

Multi-label Ranking with LSTM² for Document Classification , booktitle =

Yan Yan and Xu-Cheng Yin and Chun Yang and Bo-Wen Zhang and Hong-Wei Hao , editor =. Multi-label Ranking with LSTM² for Document Classification , booktitle =. 2016 , url =

work page 2016

[19] [19]

Transactions on Machine Learning Research , issn=

Multi-label Node Classification On Graph-Structured Data , author=. Transactions on Machine Learning Research , issn=. 2023 , url=

work page 2023

[20] [20]

Open Graph Benchmark: Datasets for Machine Learning on Graphs , booktitle =

Weihua Hu and Matthias Fey and Marinka Zitnik and Yuxiao Dong and Hongyu Ren and Bowen Liu and Michele Catasta and Jure Leskovec , editor =. Open Graph Benchmark: Datasets for Machine Learning on Graphs , booktitle =. 2020 , url =

work page 2020

[21] [21]

Sheng , title =

Cangqi Zhou and Hui Chen and Jing Zhang and Qianmu Li and Dianming Hu and Victor S. Sheng , title =. Expert Syst. Appl. , volume =. 2021 , url =

work page 2021

[22] [22]

Kipf and Max Welling , title =

Thomas N. Kipf and Max Welling , title =. CoRR , volume =. 2016 , url =

work page 2016

[23] [23]

Graph Attention Networks , booktitle =

Petar Velickovic and Guillem Cucurull and Arantxa Casanova and Adriana Romero and Pietro Li. Graph Attention Networks , booktitle =. 2018 , url =

work page 2018

[24] [24]

node2vec: Scalable Feature Learning for Networks , booktitle =

Aditya Grover and Jure Leskovec , editor =. node2vec: Scalable Feature Learning for Networks , booktitle =. 2016 , url =

work page 2016

[25] [25]

BMC Bioinform

Guangyu Cui and Chao Fang and Kyungsook Han , title =. BMC Bioinform. , volume =. 2012 , url =

work page 2012

[26] [26]

Correlation-Aware Graph Convolutional Networks for Multi-Label Node Classification , booktitle =

Yuanchen Bei and Weizhi Chen and Hao Chen and Sheng Zhou and Carl Ji Yang and Jiapei Fan and Longtao Huang and Jiajun Bu , editor =. Correlation-Aware Graph Convolutional Networks for Multi-Label Node Classification , booktitle =. 2025 , url =

work page 2025

[27] [27]

Zhi-Peng Li and Siguo Wang and Qinhu Zhang and Yi-Jie Pan and Naian Xiao and Jiayang Guo and Chang-An Yuan and Wen-Jian Liu and De-Shuang Huang , title =. Artif. Intell. Rev. , volume =. 2025 , url =

work page 2025

[28] [28]

Dhillon and Yuqiang Guan and Brian Kulis , title =

Inderjit S. Dhillon and Yuqiang Guan and Brian Kulis , title =. IEEE Trans. Pattern Anal. Mach. Intell. , volume =. 2007 , url =

work page 2007

[29] [29]

CoRR , volume =

Mikael Henaff and Joan Bruna and Yann LeCun , title =. CoRR , volume =. 2015 , url =

work page 2015

[30] [30]

Self-Attention Graph Pooling , booktitle =

Junhyun Lee and Inyeop Lee and Jaewoo Kang , editor =. Self-Attention Graph Pooling , booktitle =. 2019 , url =

work page 2019

[31] [31]

Graph U-Nets , booktitle =

Hongyang Gao and Shuiwang Ji , editor =. Graph U-Nets , booktitle =. 2019 , url =

work page 2019

[32] [32]

Structure-Feature based Graph Self-adaptive Pooling , booktitle =

Liang Zhang and Xudong Wang and Hongsheng Li and Guangming Zhu and Peiyi Shen and Ping Li and Xiaoyuan Lu and Syed Afaq Ali Shah and Mohammed Bennamoun , editor =. Structure-Feature based Graph Self-adaptive Pooling , booktitle =. 2020 , url =

work page 2020

[33] [33]

Deep Convolutional Ranking for Multilabel Image Annotation , booktitle =

Yunchao Gong and Yangqing Jia and Thomas Leung and Alexander Toshev and Sergey Ioffe , editor =. Deep Convolutional Ranking for Multilabel Image Annotation , booktitle =. 2014 , url =

work page 2014

[34] [34]

2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016 , pages =

Jiang Wang and Yi Yang and Junhua Mao and Zhiheng Huang and Chang Huang and Wei Xu , title =. 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016 , pages =. 2016 , url =

work page 2016

[35] [35]

The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025 , publisher =

Yifei Sun and Zemin Liu and Bryan Hooi and Yang Yang and Rizal Fathony and Jia Chen and Bingsheng He , title =. The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025 , publisher =. 2025 , url =

work page 2025

[36] [36]

Zheng and Kevin Chen-Chuan Chang , title =

Hongyun Cai and Vincent W. Zheng and Kevin Chen-Chuan Chang , title =. 2018 , url =

work page 2018

[37] [37]

Shunxin Xiao and Shiping Wang and Yuanfei Dai and Wenzhong Guo , title =. Mach. Vis. Appl. , volume =. 2022 , url =

work page 2022

[38] [38]

Attentive Recurrent Social Recommendation , booktitle =

Peijie Sun and Le Wu and Meng Wang , editor =. Attentive Recurrent Social Recommendation , booktitle =. 2018 , url =

work page 2018

[39] [39]

Prasanna , title =

Hanqing Zeng and Hongkuan Zhou and Ajitesh Srivastava and Rajgopal Kannan and Viktor K. Prasanna , title =. 8th International Conference on Learning Representations,. 2020 , url =

work page 2020

[40] [40]

2019 , url =

Thibaut Durand and Nazanin Mehrasa and Greg Mori , title =. 2019 , url =

work page 2019

[41] [41]

Lin Xiao and Pengyu Xu and Liping Jing and Uchenna Akujuobi and Xiangliang Zhang , title =. Inf. Sci. , volume =. 2022 , url =

work page 2022

[42] [42]

Semi-supervised Graph Embedding for Multi-label Graph Node Classification , booktitle =

Kaisheng Gao and Jing Zhang and Cangqi Zhou , editor =. Semi-supervised Graph Embedding for Multi-label Graph Node Classification , booktitle =. 2019 , url =

work page 2019

[43] [43]

2022 , doi =

Ziwei Zhang and Peng Cui and Wenwu Zhu , title =. 2022 , doi =

work page 2022

[44] [44]

Hamilton and Rex Ying and Jure Leskovec , title =

William L. Hamilton and Rex Ying and Jure Leskovec , title =. 2018 , eprint =

work page 2018

[45] [45]

2019 , eprint =

Keyulu Xu and Weihua Hu and Jure Leskovec and Stefanie Jegelka , title =. 2019 , eprint =

work page 2019

[46] [46]

Hamilton and Jure Leskovec , title =

Rex Ying and Jiaxuan You and Christopher Morris and Xiang Ren and William L. Hamilton and Jure Leskovec , title =. 2019 , eprint =

work page 2019

[47] [47]

Taylor and Mohamed R

Boris Knyazev and Graham W. Taylor and Mohamed R. Amer , title =. Proceedings of the 33rd International Conference on Neural Information Processing Systems , publisher =. 2019 , articleno =

work page 2019

[48] [48]

, title =

Zhang, Jiawei and Yu, Philip S. , title =. 2018 , issue_date =. doi:10.1145/3229329.3229333 , journal =

work page doi:10.1145/3229329.3229333 2018

[49] [49]

Hierarchical graph learning for protein–protein interaction , volume =

Gao, Ziqi and Jiang, Chenran and Zhang, Jiawen and Jiang, Xiaosen and Li, Lanqing and Zhao, Peilin and Yang, Huanming and Huang, Yong and Li, Jia , year =. Hierarchical graph learning for protein–protein interaction , volume =. Nature Communications , doi =

work page

[50] [50]

Semi-Supervised Learning Literature Survey , volume =

Zhu, Xiaojin , year =. Semi-Supervised Learning Literature Survey , volume =

work page

[51] [51]

Wu, Zonghan and Pan, Shirui and Chen, Fengwen and Long, Guodong and Zhang, Chengqi and Yu, Philip S. , year=. A Comprehensive Survey on Graph Neural Networks , volume=. IEEE Transactions on Neural Networks and Learning Systems , publisher=. doi:10.1109/tnnls.2020.2978386 , number=

work page doi:10.1109/tnnls.2020.2978386 2020

[52] [52]

2022 , eprint=

Benchmarking Graph Neural Networks , author=. 2022 , eprint=

work page 2022

[53] [53]

2024 , eprint=

Classic GNNs are Strong Baselines: Reassessing GNNs for Node Classification , author=. 2024 , eprint=

work page 2024

[54] [54]

2021 , eprint=

Bag of Tricks for Node Classification with Graph Neural Networks , author=. 2021 , eprint=

work page 2021

[55] [55]

International Conference on Learning Representations , year=

Simple Spectral Graph Convolution , author=. International Conference on Learning Representations , year=

work page

[56] [56]

2020 , eprint=

Simple and Deep Graph Convolutional Networks , author=. 2020 , eprint=

work page 2020

[57] [57]

2016 , eprint=

Layer Normalization , author=. 2016 , eprint=

work page 2016

[58] [58]

2015 , eprint=

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , author=. 2015 , eprint=

work page 2015

[59] [59]

Srivastava, Nitish and Hinton, Geoffrey and Krizhevsky, Alex and Sutskever, Ilya and Salakhutdinov, Ruslan , title =. J. Mach. Learn. Res. , month = jan, pages =. 2014 , issue_date =

work page 2014

[60] [60]

2015 , eprint=

Deep Residual Learning for Image Recognition , author=. 2015 , eprint=

work page 2015

[61] [61]

InProceedings of the 30th ACM International Conference on Information & Knowledge Management (CIKM ’21)

Song, Zixing and Meng, Ziqiao and Zhang, Yifei and King, Irwin , title =. 2021 , isbn =. doi:10.1145/3459637.3482391 , booktitle =

work page doi:10.1145/3459637.3482391 2021

[62] [62]

A Review on Multi-Label Learning Algorithms , year=

Zhang, Min-Ling and Zhou, Zhi-Hua , journal=. A Review on Multi-Label Learning Algorithms , year=

work page

[63] [63]

2020 , eprint=

Graph-Revised Convolutional Network , author=. 2020 , eprint=

work page 2020

[64] [64]

Zaki , editor =

Yu Chen and Lingfei Wu and Mohammed J. Zaki , editor =. Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings , booktitle =. 2020 , url =

work page 2020

[65] [65]

Towards Unsupervised Deep Graph Structure Learning , booktitle =

Yixin Liu and Yu Zheng and Daokun Zhang and Hongxu Chen and Hao Peng and Shirui Pan , editor =. Towards Unsupervised Deep Graph Structure Learning , booktitle =. 2022 , url =. doi:10.1145/3485447.3512186 , timestamp =

work page doi:10.1145/3485447.3512186 2022

[66] [66]

2022 , url =

Hongwei Wang and Jure Leskovec , title =. 2022 , url =. doi:10.1145/3490478 , timestamp =

work page doi:10.1145/3490478 2022

[67] [67]

Kingma and Jimmy Ba , editor =

Diederik P. Kingma and Jimmy Ba , editor =. Adam:. 3rd International Conference on Learning Representations,. 2015 , url =

work page 2015