Pith. sign in

REVIEW 4 major objections 6 minor 55 references

Disentangled Graph Representation Based on Substructure-Aware Graph Optimal Matching Kernel Convolutional Networks

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that aligning subtrees by optimal matching makes graph kernels differentiable enough to act as convolutional operators, so a GNN can learn interpretable structural patterns as trainable filters.

desk verdict A useful, moderately novel learnable graph kernel with a fixable theoretical gap: the greedy matching is not proven to yield a valid kernel, and the disentanglement claims outrun the experiments. read the letter →

arxiv 2504.16360 v1 pith:TQYDYJ2I submitted 2025-04-23 cs.LG

Mao Wang , Tao Wu , Xingping Xian , Shaojie Qiao , Weina Niu , Canyixing Cui This is my paper · ORCID
classification cs.LG
keywords graphoptimalmatchingkerneldisentangledrepresentationkernelspatternmininginterpretableneuralnetworkssubtreeencodingconvolutional
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes GOMKCN, a graph neural network that uses a graph kernel, GOMK, as its convolutional operator. For each node, GOMK compares the node-centered subgraph with several trainable graph filters by decomposing both into rooted subtrees and aligning those subtrees through matching; the matched-pair similarities form the node's disentangled representation. The authors claim this design makes graph kernels differentiable enough for end-to-end training while retaining accurate structural matching, and that the learned filters converge to the structural patterns that drive predictions. Synthetic experiments show the filters recover planted motifs, and on node and graph classification benchmarks GOMKCN matches or exceeds standard GNN baselines. If the claims hold, GNNs gain a way to learn task-relevant structural motifs as explicit, inspectable filters rather than leaving structure implicit in aggregated features.

What carries the argument

The central objects are the t-level Subtree Encoding (t-SE) and the Optimal Matching Kernel (OMK), which together form GOMK. A subgraph or filter is decomposed into subtrees rooted at each node; t-SE encodes each subtree by stacking the embeddings obtained from t rounds of weighted neighbor aggregation, so the whole subgraph becomes a set of subtree embedding vectors. OMK then measures set similarity by greedily matching each subtree in the smaller set to its most similar unmatched counterpart in the larger set and summing only matched-pair similarities, with the per-pair similarity given by a sum of RBF comparisons across levels. The theory section constructs an element kernel whose feature map is explicit via a hierarchical tree and uses the histogram intersection kernel to justify set-level similarity, which is what licenses viewing the kernel as an inner product in Hilbert space. This machinery carries the argument because it makes the graph kernel differentiable in the filter parameters while keeping the match-based accuracy that R-convolution-style sum-over-all-pairs kernels lack.

What would settle it

Run GOMK on many pairs of small random subgraphs, computing the similarity in both directions and assembling the full pairwise similarity matrix; if any pair gives different values depending on direction, or if the matrix has a negative eigenvalue, the kernel property stated in the paper fails. Repeating the test with exact Hungarian matching rather than the greedy algorithm would isolate whether the greedy step is the cause.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that graph kernels can be turned into convolutional operators without sacrificing matching accuracy, resolving what it calls the contradiction between differentiability and accuracy in existing graph kernels. The key move is to represent each node-centered subgraph and each trainable filter as a set of points in Hilbert space, with each point encoding a t-level subtree by stacking the outcomes of t rounds of neighbor aggregation. Similarity between a subgraph and a filter is then computed not by summing over all pairs of subtrees, which the paper argues injects spurious similarities, but by first aligning subtrees through an optimal matching and summing only the matched pairs. Because the filter's adjacency and feature matrices are learnable, gradient descent drives filters toward the structural patterns that matter for the task at hand. The paper also shows all subgraphs and filters with the same node count lie on a common sphere in this Hilbert space, so the kernel response is a cosine similarity, and it reads the disentangled representation of a node as the projection of its subgraph onto each learned filter.

Load-bearing premise

The paper assumes that the greedy sequential subtree matching still yields a valid kernel, even though the matching depends on which graph is the source and no symmetry or positive-semidefiniteness proof is given for the greedy version.

Editorial extensions

If this is right

  • A GNN built on GOMK can discover task-relevant structural motifs end-to-end instead of relying on a fixed, pre-enumerated motif vocabulary.
  • Because GOMK aligns subtrees before scoring, it avoids the spurious similarities that the paper attributes to summing over all substructure pairs in R-convolution kernels.
  • Node representations produced by GOMKCN are interpretable in the literal sense that a high coordinate value points to a local subgraph resembling the corresponding learned filter.
  • The spherical geometry of GOMK's feature space means kernel responses can be read as cosine similarities, so the same model can double as a frequent-subgraph miner by treating filters as cluster centers.
  • On the evaluated node and graph classification benchmarks, GOMKCN is competitive with or better than standard GNNs and the closest kernel-convolution baselines.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: because the greedy matching in Algorithm 1 is not symmetric, one should test whether the Gram matrix of GOMK is positive semidefinite on random subgraph sets; if it is not, the 'kernel' claim would need to be weakened to 'differentiable similarity layer' while most empirical uses would survive.
  • Editorial inference: the full-rank condition on t-SE embeddings suggests low-dimensional or constant node features could collapse distinct subgraphs; augmenting features or using higher-dimensional filters should make the injectivity claim hold more often in practice.
  • Editorial inference: the same formulation could be pointed at graph clustering or anomaly detection by replacing the classification loss with a kernel k-means objective, since filters already act as cluster centers in the spherical feature space.
  • Editorial inference: the filter convergence experiments leave open whether the learned patterns are stable across random initializations; a repeated-seed study on the synthetic motif datasets would show which discovered motifs are robust.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes GOMKCN, a graph convolutional network in which each node's representation is the vector of similarities between its k-hop subgraph and T learnable graph filters. The similarity is computed by decomposing subgraphs and filters into t-level subtree embeddings (t-SE) and then aligning those embeddings through a proposed Graph Optimal Matching Kernel (GOMK). The filters are trained end-to-end, and the resulting per-filter similarities are presented as disentangled structural factors. The authors claim a Hilbert-space interpretation of GOMK, a spherical feature-space property, an injectivity guarantee for t-SE, and improved interpretability and accuracy. Experiments cover isomorphic graph learning, frequent subgraph pattern mining, interpretability visualization, node classification on six datasets, and graph classification on six datasets.

Significance. If the theoretical claims held, the paper would make a useful contribution: an interpretable GNN with learnable structural filters and a differentiable graph-kernel-based convolution. The synthetic pattern-mining experiments are a concrete strength: the recovered filters in Figs. 8 and 9 provide visible evidence of interpretability, and the hyperparameter and runtime analyses are useful. The method also performs competitively, with the largest gains on ENZYMES, Chameleon, and Squirrel. However, the theoretical core is not established as written: the implemented greedy matching is not shown to define a kernel, and the proofs of element-kernel validity and t-SE injectivity have gaps. At this stage the paper is better described as an empirically motivated similarity-layer architecture than as a validated kernel-theoretic framework.

major comments (4)
  1. [Section IV-C, Eq. (16), Algorithm 1 (lines 20-25)] The kernel-valued claim is not established for the implemented function. Eq. (16) defines GOMK as an exact maximum over all matchings, but Algorithm 1 computes a greedy matching: each source subtree sequentially selects the most similar unmatched target subtree. This greedy outcome depends on the order in which source subtrees are enumerated and on which side is the source. For example, with X={x1,x2}, Y={y1,y2} and s(x1,y1)=1.0, s(x1,y2)=0.9, s(x2,y1)=0.9, s(x2,y2)=0.0 (self-similarities all 1.0), greedy from X in order x1,x2 gives total 1.0, while greedy from Y in order y2,y1 gives total 1.8. Hence the implemented kappa is not symmetric and a Gram matrix built from it need not be positive semidefinite. Proposition 5 proves validity only for a fixed matching scheme; the sentence in Section IV-C that greediness does not compromise the effectiveness of OMK to serve as a kernel function is an assertion, not a proof. The Hilbert-space interpretation (Eq. 2) and the spherical-feature-space interpretation (Eq. 21) are therefore unsupported for the model actually trained. The authors should either prove symmetry and PSD for the greedy implementation (or symmetrize it), or explicitly present GOMK as a non-kernel differentiable similarity layer and move the kernel-based theoretical claims to the exact-matching setting.
  2. [Section V-A, Proposition 6] The injectivity proof has a gap. The theorem requires the stacked embeddings hat F for G_u and G_v to be equal and full-rank, but the proof then needs each block c_i = A_G^i F to have full row rank in order to solve A_G c_i = A_H c_i uniquely for the adjacency matrix. Full rank of the stacked matrix does not imply full row rank of the individual blocks c_i; when d < n, no c_i can have rank n at all. In particular, Section VI-A reports successful recovery for featureless graphs with all node features initialized to 1.0, where F has rank 1 even though n=6. The proposition's condition is therefore not met in the experimental setting used to claim injectivity. The authors should either prove the statement under a weaker rank condition (e.g., full row rank of F), or substantially soften the expressiveness claim.
  3. [Section V-B, Eqs. (19)-(21)] Padding subgraphs with isolated zero-feature nodes is not a similarity-neutral operation. Each padded zero node contributes t+1 to the self-similarity in Eq. (19), and because RBF similarities of a zero vector to any nonzero embedding are positive (exp(-||f||^2/(d tau)) > 0), zero nodes also participate nontrivially in the greedy matching against non-isolated subtrees. The asserted bijection between the original and adjusted subgraph only preserves uniqueness of the subgraph representation; it does not preserve GOMK values. Consequently, the constant-radius sphere claim and Eq. (21)'s cosine interpretation do not follow for subgraphs of different original sizes unless the padding effect is included in the norm and the matching is made symmetric. This matters for the frequent-pattern-mining objective Eq. (23), which relies on the spherical/cosine interpretation.
  4. [Section IV-C, Propositions 5 and Eqs. (13)-(15)] Even for the exact-matching version, the proof that OMK is a kernel is incomplete. In Proposition 5 the element kernel k_e is defined only for the union of a particular pair, and the feature map psi in Eq. (13) is built from the hierarchical tree T obtained from the matching between X and Y. Therefore the set embeddings Delta_X and Delta_Y in Eq. (14) are functions of the pair (X,Y), not representations of X and Y independently. Applying the histogram intersection kernel to these pair-dependent histograms does not define a fixed positive-definite kernel on the space of all finite subsets, because there is no single feature map on each set that yields Eq. (15). To substantiate the Hilbert-space claim, the authors need a global feature map over all possible elements and a pair-independent representation of sets, or they should cite and apply the known theory of valid optimal assignment kernels with a strong base kernel.
minor comments (6)
  1. [Section V-A, Proposition 6 statement] There is a typo in the statement: it says G_kv and G_kv are isomorphic; the second occurrence should presumably be G_ku.
  2. [Section VI-E, Results] The results paragraph refers to Table II for the graph-classification results; it should refer to Table III.
  3. [Algorithm 1, line 24] The notation 'Vhat | = vhat' is not standard; it should be written as set union, for example 'Vhat <- Vhat union {vhat}'.
  4. [Section IV-C, differentiability claim] The hard argmax in Algorithm 1 is piecewise constant, so the claim of differentiability should be qualified as a subgradient statement or supported by a continuous relaxation; otherwise the phrase 'resolving the contradiction between differentiability and accuracy' overstates the property of the implemented GOMK.
  5. [Eq. (10)] The notation 1/(d tau) is unexplained: the reader cannot tell whether d is the node-feature dimension and whether the scaling is intentional; please clarify.
  6. [Table II and Table III] The phrase 'significantly higher' in Section VI-D is not backed by a statistical significance test; the standard deviations overlap for several datasets, so please either add significance tests or soften the wording.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the kernel definition, filter training, and empirical evaluation are self-contained; the greedy matching concern is a soundness gap, not a circular derivation.

full rationale

The paper's derivation chain is self-contained. The GOMK kernel value is defined directly from t-SE subtree embeddings and a matching rule (Eq. 16), and the filters are trained parameters optimized by gradient descent; no fitted parameter is renamed as a prediction. The synthetic motif experiments are supervised fitting demonstrations, not external predictions, and the real-world evaluations use held-out test sets, so there is no fit-to-prediction circularity. The theoretical claims are internal mathematical statements: Proposition 5 proves that the element kernel is PSD given a solid function and a matching scheme, Proposition 6 analyzes injectivity under full-rank conditions, and Eq. 19 follows from the definition of the solid function when exact self-matching occurs. The cited external results, such as the R-convolution framework, the histogram intersection kernel, the Moore-Aronszajn theorem, and the optimal assignment kernel proof, are used as standard support rather than as a substitute for the paper's own derivation. The only substantive weakness is that Algorithm 1's greedy matching is order-dependent and not shown to define a symmetric PSD kernel, while Section IV-C asserts that greediness 'does not compromise the effectiveness of OMK to serve as a kernel function.' That is a soundness or rigor gap about the implementation's theoretical status, not a circularity of the claimed derivation, because the formal kernel definition in Eq. 16 is independent of the greedy approximation. Overall, no load-bearing step reduces to its own inputs by construction, so the circularity score is 0.

Assumptions & free parameters 6 free parameters · 6 assumptions · 2 invented entities

The method rests on several modeling choices: the structural-factor view of graphs, learnable filters as reference patterns, and the greedy matching kernel. The free parameters are hyperparameters and learned filter weights; none are externally calibrated constants. The main unproved premise is the kernel validity of greedy matching, which is asserted without derivation. The full-rank condition in the expressivity analysis is another assumption that the paper only argues probabilistically.

free parameters (6)
  • number of graph filters T = 3 to 16 depending on dataset
    Dimensionality of the disentangled representation; tuned per dataset (Tables V and VII).
  • nodes per graph filter = 4 to 32
    Filter size matched to subgraph size; tuned per dataset.
  • subgraph radius k = 1 to 2 hops, 3 in pattern mining
    Defines the node-centric subgraph extent; hyperparameter chosen per task.
  • subtree level t = 0 to 4
    Number of aggregation iterations; performance-sensitive hyperparameter analyzed in Fig 11.
  • RBF width parameter tau = 0.5, 0.6, or 1.0
    Controls matching sharpness; fixed after model selection (Tables V and VII).
  • graph filter adjacency and feature parameters = learned by gradient descent
    These trainable parameters define the structural patterns; they are fitted to each dataset, so filter responses encode training data statistics.
assumptions (6)
  • domain assumption Almost any prediction task can be translated into recognizing and counting specific structural factors.
    Stated in the Introduction and Section IV-A; it justifies treating graphs as sets of node-centric subgraphs and is not proven.
  • ad hoc to paper The greedy matching preserves symmetry and positive semi-definiteness of the kernel.
    Asserted in Section IV-C after defining greedy matching; no proof is given and the greedy order dependence makes symmetry doubtful.
  • domain assumption The t-SE embedding matrices satisfy the full-rank condition needed for injectivity.
    Proposition 6 requires both embedding matrices to be full-rank; the paper only argues that larger t and d increase the likelihood, not a guarantee.
  • ad hoc to paper Padding subgraphs with isolated zero-feature nodes does not distort the kernel comparison.
    Introduced in Section V-B to place all subgraphs on a common sphere; zero subtrees still contribute to self-similarity and can occupy matched slots, so the operation can change kernel values.
  • standard math RBF kernels, histogram intersection kernels, and the Moore-Aronszajn theorem provide valid kernel building blocks.
    Used in Proposition 5 and the set similarity derivation; these are standard results cited as [45], [46].
  • domain assumption Node features and structural information relevant to the task are contained in k-hop subgraphs.
    The model discards information beyond k hops; the paper does not quantify the information loss for the real datasets.
invented entities (2)
  • Learnable graph filters
    purpose: Define a bank of reference structural patterns; the node representation is the vector of kernel similarities between its subgraph and each filter.
    These are model parameters, not independently observed entities. The paper offers in-paper visual checks on synthetic motifs, but no external falsifiable handle such as a new physical quantity.
  • Node-centric subgraphs as structural factors
    purpose: Treat each node-centric subgraph as a factor whose projection onto filters forms the disentangled representation.
    The independence of these factors is assumed, not measured; no metric such as mutual information is computed. The term is used to frame the method.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Disentangled Graph Representation Based on Substructure-Aware Graph Optimal Matching Kernel Convolutional Networks." pith.science (2026). https://pith.science/paper/TQYDYJ2I

@misc{pith2026250416360,
  author       = {Pith},
  title        = {Pith review of: Disentangled Graph Representation Based on Substructure-Aware Graph Optimal Matching Kernel Convolutional Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TQYDYJ2I}},
  note         = {Machine review of arXiv:2504.16360}
}
read the original abstract

Graphs effectively characterize relational data, driving graph representation learning methods that uncover underlying predictive information. As state-of-the-art approaches, Graph Neural Networks (GNNs) enable end-to-end learning for diverse tasks. Recent disentangled graph representation learning enhances interpretability by decoupling independent factors in graph data. However, existing methods often implicitly and coarsely characterize graph structures, limiting structural pattern analysis within the graph. This paper proposes the Graph Optimal Matching Kernel Convolutional Network (GOMKCN) to address this limitation. We view graphs as node-centric subgraphs, where each subgraph acts as a structural factor encoding position-specific information. This transforms graph prediction into structural pattern recognition. Inspired by CNNs, GOMKCN introduces the Graph Optimal Matching Kernel (GOMK) as a convolutional operator, computing similarities between subgraphs and learnable graph filters. Mathematically, GOMK maps subgraphs and filters into a Hilbert space, representing graphs as point sets. Disentangled representations emerge from projecting subgraphs onto task-optimized filters, which adaptively capture relevant structural patterns via gradient descent. Crucially, GOMK incorporates local correspondences in similarity measurement, resolving the trade-off between differentiability and accuracy in graph kernels. Experiments validate that GOMKCN achieves superior accuracy and interpretability in graph pattern mining and prediction. The framework advances the theoretical foundation for disentangled graph representation learning.

Figures

Figures reproduced from arXiv: 2504.16360 by the authors.

Figure 1
Figure 1. The idea of decoupled representation learning. A graph can be [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Key mechanisms in graph kernel. (a) The R-Convolution framework calculates the graph similarity by the sum of similarities between all￾substructure pairs. However, what matters may be the similarities between matched substructures, denoted by solid lines, with the dashed lines represent￾ing the spurious similarities. (b) The Label Refinement Mechanism seeks to encode a tree by mapping all node labels to a new label … view at source ↗
Figure 3
Figure 3. The framework of the Graph Optimal Matching Kernel Convolutional Network. For an input graph, the subgraphs centered around each node within [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 5
Figure 5. Figure 5: Subtree Matching based on greedy algorithm. Firstly, the similarity [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: An example of the Element Kernel ke, the elements a, b are members of set A, whereas x, y, z belong to set B. In particular, a and b are paired with x and y, respectively, with z remaining unmatched by any element. (a) The Gram matrix S describes ke with the items repr…
Figure 7
Figure 7. Figure 7: The framework and outcomes of isomorphic graph learning. (a) The framework of isomorphic graph learning. The ground truth and the graph filter [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: Illustration of frequent graph pattern mining. (a) Synthetic graph [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: Illustration of model interpretability. The first row displays four graph samples in the test set, each belonging to a distinct category. Each graph [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: Parameter sensitivity of subgraph size on Cora (top) and Citeseer [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 12
Figure 12. Figure 12: One-epoch running time of models on the ENZYMES and the [PITH_FULL_IMAGE:figures/full_fig_p014_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 31 canonical work pages

  1. [1]

    Some basic definitions in graph theory,

    J. W. Essam and M. E. Fisher, “Some basic definitions in graph theory,” Reviews of Modern Physics , vol. 42, no. 2, p. 271, 1970

    work page 1970

  2. [2]

    W. L. Hamilton, Graph representation learning . Morgan & Claypool Publishers, 2020

    work page 2020

  3. [3]

    Network representation learning: A survey,

    D. Zhang, J. Yin, X. Zhu, and C. Zhang, “Network representation learning: A survey,” IEEE transactions on Big Data , vol. 6, no. 1, pp. 3–28, 2018

    work page 2018

  4. [4]

    Deepwalk: Online learning of social representations,

    B. Perozzi, R. Al-Rfou, and S. Skiena, “Deepwalk: Online learning of social representations,” in Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining , 2014, pp. 701–710

    2014

  5. [5]

    Line: Large-scale information network embedding,

    J. Tang, M. Qu, M. Wang, M. Zhang, J. Yan, and Q. Mei, “Line: Large-scale information network embedding,” in Proceedings of the 24th international conference on world wide web , 2015, pp. 1067–1077

    2015

  6. [6]

    node2vec: Scalable feature learning for networks,

    A. Grover and J. Leskovec, “node2vec: Scalable feature learning for networks,” in Proceedings of the 22nd ACM SIGKDD international conference on Knowledge discovery and data mining , 2016, pp. 855– 864

    2016

  7. [7]

    Neural message passing for quantum chemistry,

    J. Gilmer, S. S. Schoenholz, P. F. Riley, O. Vinyals, and G. E. Dahl, “Neural message passing for quantum chemistry,” in International conference on machine learning . PMLR, 2017, pp. 1263–1272

    2017

  8. [8]

    Fraud detec- tion: A systematic literature review of graph-based anomaly detection approaches,

    T. Pourhabibi, K.-L. Ong, B. H. Kam, and Y . L. Boo, “Fraud detec- tion: A systematic literature review of graph-based anomaly detection approaches,” Decision Support Systems , vol. 133, p. 113303, 2020

    work page 2020

Show all 55 references
  1. [9]

    Drug similarity integra- tion through attentive multi-view graph auto-encoders,

    T. Ma, C. Xiao, J. Zhou, and F. Wang, “Drug similarity integra- tion through attentive multi-view graph auto-encoders,” arXiv preprint arXiv:1804.10850, 2018

    2018 arXiv

  2. [10]

    Spatial-temporal fusion graph neural networks for traffic flow forecasting,

    M. Li and Z. Zhu, “Spatial-temporal fusion graph neural networks for traffic flow forecasting,” in Proceedings of the AAAI conference on artificial intelligence, vol. 35, no. 5, 2021, pp. 4189–4196

    2021

  3. [11]

    Semi-supervised classification with graph convolutional networks,

    T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” arXiv preprint arXiv:1609.02907 , 2016

    2016 arXiv

  4. [12]

    Graph attention networks,

    P. Veli ˇckovi´c, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y . Ben- gio, “Graph attention networks,” arXiv preprint arXiv:1710.10903, 2017

    2017 arXiv

  5. [13]

    Inductive representation learning on large graphs,

    W. Hamilton, Z. Ying, and J. Leskovec, “Inductive representation learning on large graphs,” Advances in neural information processing systems, vol. 30, 2017

    2017

  6. [14]

    Node, motif and subgraph: Leveraging network functional blocks through structural convolution,

    C. Yang, M. Liu, V . W. Zheng, and J. Han, “Node, motif and subgraph: Leveraging network functional blocks through structural convolution,” in 2018 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) . IEEE, 2018, pp. 47–52

    2018

  7. [15]

    Representation learning: A review and new perspectives,

    Y . Bengio, A. Courville, and P. Vincent, “Representation learning: A review and new perspectives,” IEEE transactions on pattern analysis and machine intelligence , vol. 35, no. 8, pp. 1798–1828, 2013

    2013

  8. [16]

    Disentangled graph convolutional networks,

    J. Ma, P. Cui, K. Kuang, X. Wang, and W. Zhu, “Disentangled graph convolutional networks,” in International conference on machine learn- ing. PMLR, 2019, pp. 4212–4221

    2019

  9. [17]

    Independence promoted graph disentangled networks,

    Y . Liu, X. Wang, S. Wu, and Z. Xiao, “Independence promoted graph disentangled networks,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, no. 04, 2020, pp. 4916–4923

    2020

  10. [18]

    Lgd-gcn: Local and global disentangled graph convolutional networks,

    J. Guo, K. Huang, X. Yi, and R. Zhang, “Lgd-gcn: Local and global disentangled graph convolutional networks,” arXiv preprint arXiv:2104.11893, 2021

    2021 arXiv

  11. [19]

    Factorizable graph convolu- tional networks,

    Y . Yang, Z. Feng, M. Song, and X. Wang, “Factorizable graph convolu- tional networks,” Advances in Neural Information Processing Systems , vol. 33, pp. 20 286–20 296, 2020

    2020

  12. [20]

    Debiasing graph neural networks via learning disentangled causal substructure,

    S. Fan, X. Wang, Y . Mo, C. Shi, and J. Tang, “Debiasing graph neural networks via learning disentangled causal substructure,” Advances in Neural Information Processing Systems , vol. 35, pp. 24 934–24 946, 2022

    2022

  13. [21]

    Kergnns: Interpretable graph neural networks with graph kernels,

    A. Feng, C. You, S. Wang, and L. Tassiulas, “Kergnns: Interpretable graph neural networks with graph kernels,” in Proceedings of the AAAI conference on artificial intelligence, vol. 36, no. 6, 2022, pp. 6614–6622

    2022

  14. [22]

    Descriptive kernel convolution network with improved random walk kernel,

    M.-C. Lee, L. Zhao, and L. Akoglu, “Descriptive kernel convolution network with improved random walk kernel,” inProceedings of the ACM on Web Conference 2024 , 2024, pp. 457–468

    2024

  15. [23]

    Network motifs: simple building blocks of complex networks,

    R. Milo, S. Shen-Orr, S. Itzkovitz, N. Kashtan, D. Chklovskii, and U. Alon, “Network motifs: simple building blocks of complex networks,” Science, vol. 298, no. 5594, pp. 824–827, 2002

    2002

  16. [24]

    An algorithm for subgraph isomorphism,

    J. R. Ullmann, “An algorithm for subgraph isomorphism,” Journal of the ACM (JACM), vol. 23, no. 1, pp. 31–42, 1976

    1976

  17. [25]

    Improving graph neural network expressivity via subgraph isomorphism count- ing,

    G. Bouritsas, F. Frasca, S. Zafeiriou, and M. M. Bronstein, “Improving graph neural network expressivity via subgraph isomorphism count- ing,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 45, no. 1, pp. 657–668, 2022

    2022

  18. [26]

    Motif graph neural network,

    X. Chen, R. Cai, Y . Fang, M. Wu, Z. Li, and Z. Hao, “Motif graph neural network,” IEEE Transactions on Neural Networks and Learning Systems, 2023

    2023

  19. [27]

    Convolution kernels on discrete structures,

    D. Haussler et al., “Convolution kernels on discrete structures,” Citeseer, Tech. Rep., 1999

    1999

  20. [28]

    Fast subtree kernels on graphs,

    N. Shervashidze and K. Borgwardt, “Fast subtree kernels on graphs,” Advances in neural information processing systems , vol. 22, 2009

    2009

  21. [29]

    A linear-time graph kernel,

    S. Hido and H. Kashima, “A linear-time graph kernel,” in 2009 Ninth IEEE International Conference on Data Mining . IEEE, 2009, pp. 179– 188

    2009

  22. [30]

    Subgraph matching kernels for attributed graphs,

    N. Kriege and P. Mutzel, “Subgraph matching kernels for attributed graphs,” arXiv preprint arXiv:1206.6483 , 2012

    2012 arXiv

  23. [31]

    Marginalized kernels between labeled graphs,

    H. Kashima, K. Tsuda, and A. Inokuchi, “Marginalized kernels between labeled graphs,” in Proceedings of the 20th international conference on machine learning (ICML-03) , 2003, pp. 321–328

    2003

  24. [32]

    Weisfeiler-lehman graph kernels

    N. Shervashidze, P. Schweitzer, E. J. Van Leeuwen, K. Mehlhorn, and K. M. Borgwardt, “Weisfeiler-lehman graph kernels.” Journal of Machine Learning Research , vol. 12, no. 9, 2011

    2011

  25. [33]

    On valid optimal as- signment kernels and applications to graph classification,

    N. M. Kriege, P.-L. Giscard, and R. Wilson, “On valid optimal as- signment kernels and applications to graph classification,” Advances in neural information processing systems , vol. 29, 2016

    2016

  26. [34]

    How powerful are graph neural networks?

    K. Xu, W. Hu, J. Leskovec, and S. Jegelka, “How powerful are graph neural networks?” arXiv preprint arXiv:1810.00826 , 2018

    2018 arXiv

  27. [35]

    Disentangled graph contrastive learning with independence promotion,

    H. Li, Z. Zhang, X. Wang, and W. Zhu, “Disentangled graph contrastive learning with independence promotion,” IEEE Transactions on Knowl- edge and Data Engineering , vol. 35, no. 8, pp. 7856–7869, 2022

    2022

  28. [36]

    Interpretable deep graph generation with node-edge co-disentanglement,

    X. Guo, L. Zhao, Z. Qin, L. Wu, A. Shehu, and Y . Ye, “Interpretable deep graph generation with node-edge co-disentanglement,” in Proceed- ings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining , 2020, pp. 1697–1707

    2020

  29. [37]

    Graph kernels based on tree patterns for molecules,

    P. Mah ´e and J.-P. Vert, “Graph kernels based on tree patterns for molecules,” Machine learning, vol. 75, no. 1, pp. 3–35, 2009

    2009

  30. [38]

    Retgk: Graph kernels based on return probabilities of random walks,

    Z. Zhang, M. Wang, Y . Xiang, Y . Huang, and A. Nehorai, “Retgk: Graph kernels based on return probabilities of random walks,” Advances in Neural Information Processing Systems , vol. 31, 2018

    2018

  31. [39]

    Shortest-path kernels on graphs,

    K. M. Borgwardt and H.-P. Kriegel, “Shortest-path kernels on graphs,” in Fifth IEEE international conference on data mining (ICDM’05) . IEEE, 2005, pp. 8–pp

    2005

  32. [40]

    On graph kernels: Hardness results and efficient alternatives,

    T. G ¨artner, P. Flach, and S. Wrobel, “On graph kernels: Hardness results and efficient alternatives,” in Learning Theory and Kernel Machines: 16th Annual Conference on Learning Theory and 7th Kernel Workshop, COLT/Kernel 2003, Washington, DC, USA, August 24-27, 2003. Pro- ce...

    2003

  33. [41]

    Extensions of marginalized graph kernels,

    P. Mah ´e, N. Ueda, T. Akutsu, J.-L. Perret, and J.-P. Vert, “Extensions of marginalized graph kernels,” in Proceedings of the twenty-first inter- national conference on Machine learning , 2004, p. 70

    2004

  34. [42]

    The expressive power of graph neural networks: A survey,

    B. Zhang, C. Fan, S. Liu, K. Huang, X. Zhao, J. Huang, and Z. Liu, “The expressive power of graph neural networks: A survey,” IEEE Transactions on Knowledge and Data Engineering , 2024

    2024

  35. [43]

    The hungarian method for the assignment problem,

    H. W. Kuhn, “The hungarian method for the assignment problem,” Naval research logistics quarterly, vol. 2, no. 1-2, pp. 83–97, 1955

    1955

  36. [44]

    Algorithms for the assignment and transportation prob- lems,

    J. Munkres, “Algorithms for the assignment and transportation prob- lems,” Journal of the society for industrial and applied mathematics , vol. 5, no. 1, pp. 32–38, 1957

    1957

  37. [45]

    Theory of reproducing kernels,

    N. Aronszajn, “Theory of reproducing kernels,” Transactions of the American mathematical society , vol. 68, no. 3, pp. 337–404, 1950

    1950

  38. [46]

    Generalized histogram intersection kernel for image recognition,

    S. Boughorbel, J.-P. Tarel, and N. Boujemaa, “Generalized histogram intersection kernel for image recognition,” in IEEE International Con- ference on Image Processing 2005 , vol. 3. IEEE, 2005, pp. III–161

    2005

  39. [47]

    Revisiting semi-supervised learning with graph embeddings,

    Z. Yang, W. Cohen, and R. Salakhudinov, “Revisiting semi-supervised learning with graph embeddings,” in International conference on ma- chine learning. PMLR, 2016, pp. 40–48

    2016

  40. [48]

    Multi-scale attributed node embedding,

    B. Rozemberczki, C. Allen, and R. Sarkar, “Multi-scale attributed node embedding,” Journal of Complex Networks , vol. 9, no. 2, p. cnab014, 2021

    2021

  41. [49]

    Geom-gcn: Geo- metric graph convolutional networks,

    H. Pei, B. Wei, K. C.-C. Chang, Y . Lei, and B. Yang, “Geom-gcn: Geo- metric graph convolutional networks,” arXiv preprint arXiv:2002.05287, 2020

    2002 arXiv

  42. [50]

    Distinguishing enzyme structures from non-enzymes without alignments,

    P. D. Dobson and A. J. Doig, “Distinguishing enzyme structures from non-enzymes without alignments,” Journal of molecular biology , vol. 330, no. 4, pp. 771–783, 2003

    2003

  43. [51]

    Brenda, the enzyme database: updates and major new developments,

    I. Schomburg, A. Chang, C. Ebeling, M. Gremse, C. Heldt, G. Huhn, and D. Schomburg, “Brenda, the enzyme database: updates and major new developments,” Nucleic acids research , vol. 32, no. suppl 1, pp. D431–D433, 2004. 16

    2004

  44. [52]

    Protein function prediction via graph ker- nels,

    K. M. Borgwardt, C. S. Ong, S. Sch ¨onauer, S. Vishwanathan, A. J. Smola, and H.-P. Kriegel, “Protein function prediction via graph ker- nels,” Bioinformatics, vol. 21, no. suppl 1, pp. i47–i56, 2005

    2005

  45. [53]

    Deep graph kernels,

    P. Yanardag and S. Vishwanathan, “Deep graph kernels,” in Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining , 2015, pp. 1365–1374

    2015

  46. [54]

    Graph u-nets,

    H. Gao and S. Ji, “Graph u-nets,” in international conference on machine learning. PMLR, 2019, pp. 2083–2092

    2019

  47. [55]

    Self-attention graph pooling,

    J. Lee, I. Lee, and J. Kang, “Self-attention graph pooling,” in Interna- tional conference on machine learning . pmlr, 2019, pp. 3734–3743. APPENDIX A TABLE IV DATASET STATISTICS FOR NODE CLASSIFICATION . Dataset Cora Citeseer Pubmed Chameleon Squirrel Actor Nodes 2708 3327 19...

    2019

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.