Pith. sign in

REVIEW 4 major objections 5 minor 105 references

Pattern-Based Graph Classification: Comparison of Quality Measures and Importance of Preprocessing

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read A comparison of 38 quality measures for pattern-based graph classification finds that AbsSupDif and Sup are the safest choices, while popular measures such as GR, Acc, and InfGain perform considerably worse.

desk verdict Useful graph-specific comparison of quality measures, but the F1-based recommendations need a documented train/test protocol before they can be trusted. read the letter →

arxiv 2507.00039 v1 pith:2DR6JKOI submitted 2025-06-19 cs.LG cs.AI

classification cs.LGcs.AI
keywords graphclassificationpatternminingqualitymeasuressubgraphpatternsShapleyvaluehierarchicalclusteringpreprocessingbinary
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

The paper asks which of the many quality measures used to rank subgraph patterns actually helps graph classification. It compares 38 measures on eight datasets, characterizes them with four mathematical properties, and builds a gold-standard pattern ranking from Shapley values. Its empirical conclusion is that AbsSupDif (the absolute difference between pattern presence in the positive and negative classes) and Sup (presence in the positive class) are safe choices across datasets, while popular measures such as GR, Acc, and InfGain are considerably less effective. The paper also proposes a clustering preprocessing step that groups patterns with similar footprints, and shows that this reduces the number of patterns while achieving comparable or better classification performance.

What carries the argument

The footprint of a pattern is the binary vector recording, for every graph in the collection, whether that pattern occurs. The paper's clustering step groups patterns whose footprints are close under Manhattan distance using complete-linkage hierarchical clustering, selects one medoid per cluster as representative, and ranks only representatives, removing patterns that are interchangeable from the classifier's perspective. The gold standard uses Shapley values, approximated globally, to score each representative's contribution to classification performance; measure rankings are compared with Kendall's Tau for pairwise equivalence and with Rank-Biased Overlap against the gold standard. Four properties—Contrastivity, Jumpiness, Class Symmetry, and Pattern Symmetry—are introduced to explain why measures differ.

What would settle it

Compute exact Shapley values for all patterns on a small dataset and compare them with the approximate gold standard; if the two rankings disagree substantially, or if AbsSupDif and Sup no longer rank near the top under the exact values, the central recommendation collapses.

Watch

Extended reading notes

Core claim

The paper's central empirical claim is that, among 38 quality measures used to rank mined subgraph patterns for binary graph classification, AbsSupDif and Sup are consistently good choices across eight datasets, while several popular measures are not. Using a gold-standard ranking built from Shapley values approximated over pattern contributions, the paper compares measures two ways: how well their rankings of cluster representatives agree with the gold standard, and how fast F1-score rises when the top-ranked patterns feed a classifier. It finds that GR, Acc, and InfGain, despite being widespread, rank patterns poorly relative to the gold standard and need many more patterns to reach similar performance. It also claims that clustering patterns by footprint distance before ranking reduces the number of patterns by up to 92% while achieving comparable or better classification performance, and that groups of measures produce identical rankings, collapsing 38 measures to 21 informative ones.

Load-bearing premise

The conclusions rely on the Shapley-value approximation being a true gold standard for which patterns are most useful for classification, even though the paper notes that one measure, Dep, beats that gold standard on IMDb.

Editorial extensions

If this is right

  • Practitioners with no prior knowledge about a graph dataset can safely pick AbsSupDif or Sup to rank patterns; the paper reports they perform well on all eight datasets.
  • Popular choices such as GR, Acc, and InfGain are not reliable defaults for pattern-based graph classification and can require many more patterns to reach the F1-score of the better measures.
  • Many of the 38 measures are redundant: six blocks of measures produce identical rankings, so future studies can restrict attention to one representative per block.
  • A footprint-based clustering preprocessing step can shrink the pattern set substantially, for example by about 92% on MUTAG, while keeping or slightly improving classification performance, making larger graph collections computationally feasible.
  • The natural next steps stated by the paper are extending the comparison to imbalanced and multiclass settings and benchmarking against methods that mine discriminative patterns directly.

Reading between the lines

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

  • Because the equivalence blocks imply that some measures are interchangeable, a testable extension is to replace an arbitrary measure with AbsSupDif and check whether classification performance and runtime improve on unseen datasets, not just the eight studied.
  • The clustering step may serve as a general denoising preprocessing for any pattern-based representation, not only quality-measure ranking; one could test it before graph-kernel or graph-neural-network input construction.
  • The gold standard is only as good as the Shapley approximation; if exact Shapley values were computed on small graphs, the ranking comparisons could be re-run to check whether the top measures remain AbsSupDif and Sup.
  • The paper's evidence that Dep beats the gold standard on IMDb suggests the approximate gold standard may be less reliable on social-network-style graphs, so the safe-choice recommendation may be strongest on molecular datasets.
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 / 5 minor

Summary. The paper presents a comparative study of 38 quality measures for pattern-based graph classification. It introduces four theoretical properties (Contrastivity, Jumpiness, Class Symmetry, Pattern Symmetry), proposes a clustering-based preprocessing step that groups patterns with similar footprints, and constructs a gold standard ranking from Shapley-value-based importance scores. The measures are evaluated on eight public graph datasets by comparing their rankings with the gold standard (using Kendall's Tau and Rank-Biased Overlap) and by measuring F1-score when the top-ranked patterns are used as features for an SVM classifier. The main claims are that AbsSupDif and Sup are safe choices across datasets, that popular measures such as GR, Acc, and InfGain are comparatively less effective, and that the clustering preprocessing reduces the number of patterns while maintaining or improving classification performance.

Significance. If the empirical claims were fully supported, the paper would provide a useful reference for practitioners choosing quality measures for subgraph-based classification, and the clustering preprocessing idea is sensible and potentially valuable. The theoretical characterization through four properties is a genuine contribution, and the public release of code, datasets, and experimental results is a strength that aids reproducibility. However, the central F1-based conclusions currently rest on an underspecified and potentially in-sample evaluation protocol, and the gold standard used for ranking comparisons is itself an approximation whose validity is not established. These issues must be resolved before the recommendations can be accepted as reliable.

major comments (4)
  1. [Section 6.1, 6.2.3, 6.4.2] No train/test split, cross-validation, repeated runs, or error bars are described for any of the F1-based evaluations. Section 6.2.3 states that the authors 'train a classifier' and 'assess its classification performance with the F1-Score', and Section 6.4.2 similarly says that the authors 'train the classifier and compute the classification performance', but neither section nor Section 6.1 specifies which graphs are used for training and which for testing. Under a literal reading, the F1 curves in Figures 8 and 11 are computed on the same graphs used for training. If so, the ordering of measures in Section 6.4.2 and the conclusion that AbsSupDif and Sup are safe choices while GR and Acc are less relevant cannot be interpreted as predictive performance. Even if the released code performs cross-validation, the manuscript must state the protocol explicitly and report variance across runs.
  2. [Section 6.2.3, 6.3, 6.4] The per-dataset clustering threshold is selected from the very same F1 curves that are later used to evaluate the clustering benefit and the quality measures. In Section 6.2.3, the vertical dotted lines in Figure 8 are chosen as the 'best trade-off' between minimizing the number of representatives and maximizing classification performance; these thresholds are then reused for the ranking comparisons in Section 6.3 and the gold standard comparison in Section 6.4. Selecting a parameter on the evaluation data creates a selection bias that can inflate the apparent benefit of clustering and can distort the subsequent comparisons between measures. A nested or independent validation scheme is needed to support the conclusions.
  3. [Section 5.2.1, Section 6.1, Appendix D.2] The gold standard used throughout Section 6.4 is described as based on the Shapley Value via SAGE, but the actual implementation uses LossSHAP, which the paper states 'only provides a local version of SAGE' and is averaged over all patterns to obtain global scores. Averaging local SHAP values over data points is not equivalent to computing SAGE values, which are defined globally with respect to the model loss. The validity of the RBO comparisons and the claim that AbsSupDif and Sup are close to the gold standard depends on this approximation. The paper itself notes in Appendix D.2 that Dep outperforms the gold standard on IMDb, indicating that the gold standard is imperfect. The authors should either compute genuine SAGE values or explicitly justify and quantify the error introduced by the local approximation.
  4. [Section 5.2.2, Eq. (3)] The Rank-Biased Overlap is defined in Eq. (3) as an infinite sum, and the text states that 'in our case' the upper bound is s, but no finite-list correction or normalization is provided. With the truncated sum, the maximum possible RBO when both rankings are identical at all depths s is 1-p^s, not 1, so the upper bound depends on s. This means the increasing RBO curves in Figure 10 may be partly a mechanical consequence of the truncation. In addition, the value of the parameter p is never specified in Section 6.1, even though it controls the top-weighting and therefore directly affects the numerical comparisons. The authors should state the chosen p, use a proper finite-list variant of RBO, and report sensitivity to p.
minor comments (5)
  1. [Section 7] The conclusion cites InfGain as originating from [85], but in Table 2 InfGain is attributed to [18]; the reference should be corrected.
  2. [Appendix B.2] The text refers to 'Figure 6.3 from Section 9' when discussing the minimum Kendall's Tau matrix; this should be Figure 9 from Section 6.3.
  3. [Section 6.2.3] The sentence 'This choice allows us to focus on the impact of the clustering process on classification, rather than on rather than on the nature of the classifier' contains a duplicated phrase and should be rewritten.
  4. [Section 7] The conclusion states that the authors 'also show empirically that restricting pattern mining to specific types of patterns, such as induced or closed ones, also results in a smaller selection of patterns for equal performance', but no such experiments are presented in Section 6 or in the appendices; this claim should be removed or supported with results.
  5. [Appendix C.2] The introductory sentence says that 'Figures 15 and 16 show the F1-Score' for each measure, but those figures display RBO values; the F1-Score plots are Figures 17 and 18 and the cross-reference should be fixed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the quality measures and the empirical comparisons are independently defined and self-contained.

full rationale

The paper's derivation chain is empirical rather than definitional. Section 4 defines 38 quality measures from support-based probability tables that are independent of the paper's own conclusions; none of the formulas is defined in terms of the gold standard or of the classification outcome. The clustering step (Section 5.1) applies hierarchical agglomerative clustering over pattern footprints, and the representative selection is a standard medoid computation; this is not a fitted quantity being renamed as a prediction. The SAGE/LossSHAP gold standard (Section 5.2.1) is an external Shapley-value-based proxy built on a prediction model, and the paper explicitly acknowledges in Appendix D.2 that it is only an approximation, since Dep outperforms it on IMDb. Any weakness there is a validity concern about the benchmark, not circular derivation. The comparisons in Section 6.4 use RBO and F1-score against this gold standard; the conclusion that AbsSupDif and Sup are safe choices is an empirical outcome, not an identity with the inputs. The only self-citation, Potin et al. [73], is used as an illustrative application and as a dataset source, not as a load-bearing mathematical premise. Concerns about the lack of a described train/test split and about tuning the clustering threshold on the evaluation datasets are legitimate correctness or selection-bias issues, but they are not instances of a claim reducing to its own inputs by construction. Consequently, no circular step is exhibited and the circularity score is 0.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central comparison rests on the SAGE gold standard, the pattern mining thresholds, the clustering threshold tuning, and the transfer of tabular measures to graph support. No new physical entities are proposed, and no fitted constants appear inside the quality measures themselves.

free parameters (3)
  • Per-dataset clustering threshold = Not numerically reported; selected from the best trade-off points in Fig. 8
    Chosen on each dataset from F1 curves and then reused for ranking and classification comparisons in Sections 6.3 and 6.4. This is a data-dependent tuning choice that can inflate apparent performance.
  • Minimum support threshold for pattern mining = 0%, 1%, 1%, 25%, 1%, 0%, 1%, 20% for MUTAG, PTC, NCI1, D&D, AIDS, FOPPA, IMDb, FRANK (Table 5)
    Chosen empirically per dataset to keep mining tractable. It directly determines the pool of patterns that are later ranked, so it is a free parameter affecting the central comparison.
  • RBO top-weight parameter p = Not stated
    Eq. 3 defines Rank-Biased Overlap with a parameter p that controls how much top patterns matter, but the paper never reports the value used in the RBO figures. This is a hidden parameter for the ranking comparisons.
assumptions (5)
  • domain assumption SAGE/LossSHAP Shapley values approximate the true discriminative contribution of each pattern to the classification task.
    Used to build the gold standard ranking in Section 5.2.1; Appendix D.2 concedes it is an approximation.
  • domain assumption The mined pattern set, after per-dataset support thresholds, is sufficiently representative for fair measure comparison.
    Table 4 notes non-exhaustive pattern search on several datasets due to computational limits.
  • domain assumption Balanced binary classes do not invalidate conclusions about quality measures.
    Assumed throughout Section 3; the authors list unbalanced and multiclass settings as future work in Section 7.
  • domain assumption Complete-linkage hierarchical clustering with Manhattan distance creates meaningful groups of interchangeable patterns.
    Section 5.1 relies on this to reduce patterns; relaxed clusters may discard discriminative patterns, so the threshold choice is load-bearing.
  • domain assumption Quality measures defined for tabular itemsets transfer to graph support in a meaningful way.
    Section 4.1 adapts tabular measures using graph support; one measure, SupMaxK, cannot be transferred, which shows the transfer is not trivial.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pattern-Based Graph Classification: Comparison of Quality Measures and Importance of Preprocessing." pith.science (2026). https://pith.science/paper/2DR6JKOI

@misc{pith2026250700039,
  author       = {Pith},
  title        = {Pith review of: Pattern-Based Graph Classification: Comparison of Quality Measures and Importance of Preprocessing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2DR6JKOI}},
  note         = {Machine review of arXiv:2507.00039}
}
read the original abstract

Graph classification aims to categorize graphs based on their structural and attribute features, with applications in diverse fields such as social network analysis and bioinformatics. Among the methods proposed to solve this task, those relying on patterns (i.e. subgraphs) provide good explainability, as the patterns used for classification can be directly interpreted. To identify meaningful patterns, a standard approach is to use a quality measure, i.e. a function that evaluates the discriminative power of each pattern. However, the literature provides tens of such measures, making it difficult to select the most appropriate for a given application. Only a handful of surveys try to provide some insight by comparing these measures, and none of them specifically focuses on graphs. This typically results in the systematic use of the most widespread measures, without thorough evaluation. To address this issue, we present a comparative analysis of 38 quality measures from the literature. We characterize them theoretically, based on four mathematical properties. We leverage publicly available datasets to constitute a benchmark, and propose a method to elaborate a gold standard ranking of the patterns. We exploit these resources to perform an empirical comparison of the measures, both in terms of pattern ranking and classification performance. Moreover, we propose a clustering-based preprocessing step, which groups patterns appearing in the same graphs to enhance classification performance. Our experimental results demonstrate the effectiveness of this step, reducing the number of patterns to be processed while achieving comparable performance. Additionally, we show that some popular measures widely used in the literature are not associated with the best results.

Figures

Figures reproduced from arXiv: 2507.00039 by the authors.

Figure 1
Figure 1. Processing steps of a standard pattern-based graph classification framework. [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Simplified representation of a collection [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Processing steps of the extended framework, including the additional clustering step (1a), in blue. [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (17 more)
Figure 4
Figure 4. Figure 4: Illustration of the issue occurring when comparing rankings of patterns possessing similar footprints in the original framework. [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: The top part of the figure positions this step in the general pipeline, whereas the bottom part provides an [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: Number of representatives as a function of the clustering threshold. Note the logarithmic scale of the [PITH_FULL_IMAGE:figures/full_fig_p022_6.png]
Figure 7
Figure 7. Figure 7: Distribution of Kendall’s Tau over all pairs of quality measures, for four values of the clustering threshold (0%, 20%, 40% and [PITH_FULL_IMAGE:figures/full_fig_p023_7.png]
Figure 8
Figure 8. Figure 8: Classification performance (𝐹 1-Score) as a function of the clustering threshold. The vertical dotted black lines materialize the threshold values used in the rest of our experiments. values (black lines) are very low for these datasets. For all datasets, we observe a …
Figure 9
Figure 9. Figure 9: Minimal value of Kendall’s Tau over all datasets, for each pair of quality measures. [PITH_FULL_IMAGE:figures/full_fig_p026_9.png]
Figure 10
Figure 10. Figure 10: RBO between the gold standard and the rankings obtained for the eight quality measure of interest, as a function of [PITH_FULL_IMAGE:figures/full_fig_p028_10.png]
Figure 11
Figure 11. Figure 11: 𝐹 1-Score as a function of the proportion of representatives selected, for each quality measure of interest, as well as the gold standard (dotted line). To visualize all quality measures, see Appendix C.2. specific ranking produced by a measure only affects which patt…
Figure 12
Figure 12. Figure 12: Distribution of Kendall’s Tau coefficient computed over all pairs of quality measure, for datasets MUTAG, PTC, and NCI1. [PITH_FULL_IMAGE:figures/full_fig_p041_12.png]
Figure 13
Figure 13. Figure 13: Distribution of Kendall’s Tau coefficient computed over all pairs of quality measure, for datasets D&D, AIDS, and FOPPA. [PITH_FULL_IMAGE:figures/full_fig_p042_13.png]
Figure 14
Figure 14. Figure 14: Kendall’s Tau for each pair of quality measures, shown separately for each dataset. Note that the color scale is not fixed over [PITH_FULL_IMAGE:figures/full_fig_p043_14.png]
Figure 15
Figure 15. Figure 15: RBO between the rankings obtained for each selected quality measure and the gold standard, as a function of [PITH_FULL_IMAGE:figures/full_fig_p044_15.png]
Figure 16
Figure 16. Figure 16: RBO between the rankings obtained for each selected quality measure and the gold standard, as a function of [PITH_FULL_IMAGE:figures/full_fig_p045_16.png]
Figure 17
Figure 17. Figure 17: 𝐹 1-Score as a function of the proportion of representatives selected for each quality measure and gold standard for datasets MUTAG, PTC and NCI1. The rest of the datasets are shown in [PITH_FULL_IMAGE:figures/full_fig_p046_17.png]
Figure 18
Figure 18. Figure 18: 𝐹 1-Score as a function of the proportion of representatives selected for each quality measure and gold standard for datasets D&D, AIDS and FOPPA. The rest of the datasets are shown in [PITH_FULL_IMAGE:figures/full_fig_p047_18.png]
Figure 19
Figure 19. Figure 19: Experiments for the FRANK dataset. The main difference between this dataset and the others, in terms of results, is in the blocks of correlation quality measures identified using Kendall’s Tau (Figure 19c), which are not exactly the same as for the other datasets: • D…
Figure 20
Figure 20. Figure 20: Experiments for the IMDb dataset. The blocks of measures are identical to the general case. However, a difference can be observed regarding classification performance. Measure Dep achieves a better 𝐹 1-score than our gold standard, despite a low RBO between the two. T…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

105 extracted references · 47 canonical work pages

  1. [1]

    Acosta-Mendoza, A

    N. Acosta-Mendoza, A. Gago-Alonso, J. A. Carrasco-Ochoa, J. Francisco Martínez-Trinidad, and J. Eladio Medina-Pagola. 2016. Improving graph-based image classification by using emerging patterns as attributes. Engineering Applications of Artificial Intelligence 50 (2016), 215–225. https://doi.org/10.1016/j.engappai.2016.01.030

  2. [2]

    C. C. Aggarwal, M. A. Bhuiyan, and M. A. Hasan. 2014. Frequent Pattern Mining Algorithms: A Survey. In Frequent Pattern Mining. Springer, Chapter 2, 19–64. https://doi.org/10.1007/978-3-319-07821-2_2 Manuscript submitted to ACM 32 *

  3. [3]

    Agrawal, T

    R. Agrawal, T. Imieliński, and A. Swami. 1993. Mining association rules between sets of items in large databases. ACM SIGMOD Record 22, 2 (1993), 207–216. https://doi.org/10.1145/170036.170072

  4. [4]

    Ahn and J

    K. Ahn and J. Kim. 2004. Efficient Mining of Frequent Itemsets and a Measure of Interest for Association Rule Mining. Journal of Information and Knowledge Management 03, 03 (2004), 245–257. https://doi.org/10.1142/s0219649204000869

  5. [5]

    M. T. Alam, C. F. Ahmed, M. Samiullah, and C. K. Leung. 2021. Discriminating Frequent Pattern Based Supervised Graph Embedding for Classification. In 25th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining (Lecture Notes in Computer Science, Vol. 12713) . Springer, 16–28. https://doi.org/10.1007/978-3-030-75765-6_2

  6. [6]

    K. Ali, S. Manganaris, and R. Srikant. 1997. Partial classification using association rules. In 3rd International Conference on Knowledge Discovery and Data Mining. 115–118. https://dl.acm.org/doi/10.5555/3001392.3001412

  7. [7]

    An and N

    A. An and N. Cercone. 1998. ELEM2: A learning system for more accurate classifications. In Conference of the Canadian Society for Computational Studies of Intelligence (Lecture Notes in Computer Science, Vol. 1418) . Springer, 426–441. https://doi.org/10.1007/3-540-64575-6_68

  8. [8]

    An and N

    A. An and N. Cercone. 1999. An Empirical Study on Rule Quality Measures. In International Workshop on Rough Sets, Fuzzy Sets, Data Mining, and Granular-Soft Computing (Lecture Notes in Computer Science, Vol. 1711) . Springer, 482–491. https://doi.org/10.1007/978-3-540-48061-7_59

Show all 105 references
  1. [9]

    An and N

    A. An and N. Cercone. 2001. Rule Quality Measures for Rule Induction Systems: Description and Evaluation. Computational Intelligence 17, 3 (2001), 409–424. https://doi.org/10.1111/0824-7935.00154

  2. [10]

    Bandyopadhyaya and R

    V. Bandyopadhyaya and R. Bandyopadhyaya. 2021. Understanding the Impact of COVID-19 Pandemic Outbreak on Grocery Stocking Behaviour in India: A Pattern Mining Approach. Global Business Review 25, 3 (Feb. 2021), 750–770. https://doi.org/10.1177/0972150921988955

  3. [11]

    M. C. Barbieri, B. I. Grisci, and M. Dorn. 2024. Analysis and comparison of feature selection methods towards performance and stability. Expert Systems with Applications 249 (2024), 123667. https://doi.org/10.1016/j.eswa.2024.123667

  4. [12]

    S. D. Bay and M. J. Pazzani. 1999. Detecting change in categorical data: mining contrast sets. In 5th ACM SIGKDD international conference on Knowledge discovery and data mining . 302–306. https://doi.org/10.1145/312129.312263

  5. [13]

    Breiman, J

    L. Breiman, J. H. Friedman, R. A. Olshen, and C. J. Stone. 1984.Classification And Regression Trees. Routledge. https://doi.org/10.1201/9781315139470

  6. [14]

    S. Brin, R. Motwani, J. D. Ullman, and S. Tsur. 1997. Dynamic itemset counting and implication rules for market basket data. ACM SIGMOD Record 26, 2 (1997), 255–264. https://doi.org/10.1145/253262.253325

  7. [15]

    C. C. Chang and C. J. Lin. 2011. LIBSVM: a library for support vector machines. ACM Transactions on Intelligent Systems and Technology 2, 3 (2011), 1–27. https://doi.org/10.1145/1961189.1961199

  8. [16]

    Y. Chen, W. Gan, Y. Wu, and P. S. Yu. 2022. Contrast Pattern Mining: A Survey. arXiv cs.DB (2022), 2209.13556. https://arxiv.org/abs/2209.13556

  9. [17]

    M. E. S. Chowdhury, C. F. Ahmed, and C. K. Leung. 2021. A New Approach for Mining Correlated Frequent Subgraphs. ACM Transactions on Management Information Systems 13, 1 (2021), 9. https://doi.org/10.1145/3473042

  10. [18]

    K. W. Church and P. Hanks. 1990. Word association norms, mutual information, and lexicography. Computational Linguistics 16, 1 (1990), 22–29. https://aclanthology.org/J90-1003

  11. [19]

    Cortes and V

    C. Cortes and V. Vapnik. 1995. Support-vector networks. Machine Learning 20, 3 (1995), 273–297. https://doi.org/10.1007/bf00994018

  12. [20]

    Covert, S

    I. Covert, S. M. Lundberg, and S.-I. Lee. 2020. Understanding Global Feature Contributions With Additive Importance Measures. In 34th Conference on Neural Information Processing Systems . 17212–17223. https://proceedings.neurips.cc/paper_files/paper/2020/hash/ c7bf0b7c1a86d5eb...

  13. [21]

    A. S. Debnath, R. L. Lopez, G. Debnath, A. Shusterman, and C. Hansch. 1991. Structure-activity relationship of mutagenic aromatic and heteroaromatic nitro compounds. Correlation with molecular orbital energies and hydrophobicity. Journal of Medicinal Chemistry 34, 2 (1991), 78...

  14. [22]

    Dhal and C

    P. Dhal and C. Azad. 2021. A comprehensive survey on feature selection in the various fields of machine learning. Applied Intelligence 52, 4 (2021), 4543–4581. https://doi.org/10.1007/s10489-021-02550-9

  15. [23]

    P. D. Dobson and A. J. Doig. 2003. Distinguishing Enzyme Structures from Non-enzymes Without Alignments. Journal of Molecular Biology 330, 4 (2003), 771–783. https://doi.org/10.1016/s0022-2836(03)00628-4

  16. [24]

    Contrast Data Mining: Concepts, Algorithms, and Applications

    2012. Contrast Data Mining: Concepts, Algorithms, and Applications . Chapman and Hall/CRC. https://doi.org/10.1201/b12986

  17. [25]

    Dong and J

    G. Dong and J. Li. 1999. Efficient mining of emerging patterns: discovering trends and differences. In 5th ACM SIGKDD international conference on Knowledge discovery and data mining . 43–52. https://doi.org/10.1145/312129.312191

  18. [26]

    Fagin, R

    R. Fagin, R. Kumar, and D. Sivakumar. 2003. Comparing Top k Lists. SIAM Journal on Discrete Mathematics 17, 1 (2003), 134–160. https: //doi.org/10.1137/s0895480102412856

  19. [27]

    G. Fang, W. Wang, B. Oatley, B. Van Ness, M. Steinbach, and V. Kumar. 2011. Characterizing Discriminative Patterns.arXiv cs.DB (2011), 1102.4104. https://arxiv.org/abs/1102.4104

  20. [28]

    R. A. Fisher. 1936. The use of multiple measurements in taxonomic problems.Annals of Eugenics 7, 2 (1936), 179–188. https://doi.org/10.1111/j.1469- 1809.1936.tb02137.x

  21. [29]

    Fournier-Viger, C

    P. Fournier-Viger, C. Cheng, J. Chun-Wei Lin, U. Yun, and R. U. Kiran. 2019. TKG: Efficient Mining of Top-K Frequent Subgraphs. InInternational Conference on Big Data Analytics (Lecture Notes in Computer Science, Vol. 11932) . Springer, 209–226. https://doi.org/10.1007/978-3-0...

  22. [30]

    Fournier-Viger, J

    P. Fournier-Viger, J. C. W. Lin, A. Gomariz, T. Gueniche, A. Soltani, Z. Deng, and H. T. Lam. 2016. The SPMF Open-Source Data Mining Library Version 2. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases (Lecture Notes in Computer Science, Vol...

  23. [31]

    Fryer, I

    D. Fryer, I. Strumke, and H. Nguyen. 2021. Shapley Values for Feature Selection: The Good, the Bad, and the Axioms. IEEE Access 9 (2021), 144352–144360. https://doi.org/10.1109/access.2021.3119110

  24. [32]

    A. M. Garcia-Vico, C. J. Carmona, P. Gonzalez, H. Seker, and M. J. Jesus. 2020. FEPDS: A Proposal for the Extraction of Fuzzy Emerging Patterns in Data Streams. IEEE Transactions on Fuzzy Systems 28, 12 (2020), 3193–3203. https://doi.org/10.1109/tfuzz.2020.2992849

  25. [33]

    García-Borroto, O

    M. García-Borroto, O. Loyola-Gonzalez, J. F. Martínez-Trinidad, and J. A. Carrasco-Ochoa. 2013. Comparing Quality Measures for Contrast Pattern Classifiers. In 18th Iberoamerican Congress on Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications (Le...

  26. [34]

    García-Borroto, J

    M. García-Borroto, J. F. Martínez-Trinidad, J. A. Carrasco-Ochoa, M. A. Medina-Pérez, and J. Ruiz-Shulcloper. 2010. LCMine: An efficient algorithm for mining discriminative regularities and its application in supervised classification. Pattern Recognition 43, 9 (2010), 3025–30...

  27. [35]

    J. E. Gentle, L. Kaufman, and P. J. Rousseuw. 1991. Finding Groups in Data: An Introduction to Cluster Analysis. Biometrics 47, 2 (1991), 788. https://doi.org/10.2307/2532178

  28. [36]

    R. Gras, S. Ag Almouloud, M. Bailleul, A. Larher, M. Polo, H. Ratsimba-Rajohn, and A. Totohasina. 1996. L’implication statistique, nouvelle méthode exploratoire de données. La Pensée Sauvage. https://cir.nii.ac.jp/crid/1130282269046733952

  29. [37]

    Guo and X

    T. Guo and X. Zhu. 2013. Understanding the roles of sub-graph features for graph classification: an empirical study perspective. In 22nd ACM international conference on information and knowledge management . ACM Press, 817–822. https://doi.org/10.1145/2505515.2505614

  30. [38]

    Güvenoglu and B

    B. Güvenoglu and B. E. Bostanoglu. 2018. A qualitative survey on frequent subgraph mining. Open Computer Science 8, 1 (2018), 194–209. https://doi.org/10.1515/comp-2018-0018

  31. [39]

    Harchaoui and F

    Z. Harchaoui and F. Bach. 2007. Image Classification with Segmentation Graph Kernels. In IEEE Conference on Computer Vision and Pattern Recognition. 1–8. https://doi.org/10.1109/cvpr.2007.383049

  32. [40]

    B. Harris. 1966. The Estimation of Probabilities: An Essay on Modern Bayesian Methods (I. J. Good). SIAM Rev. 8, 1 (1966), 118–119. https: //doi.org/10.1137/1008024

  33. [41]

    C. He, X. Chen, G. Chen, W. Gan, and P.Ò Fournier-Viger. 2024. Mining credible attribute rules in dynamic attributed graphs. Expert Systems with Applications 246 (2024), 123012. https://doi.org/10.1016/j.eswa.2023.123012

  34. [42]

    Hellal and L

    A. Hellal and L. Ben Romdhane. 2016. Minimal contrast frequent pattern mining for malware detection. Computers & Security 62 (2016), 19–32. https://doi.org/10.1016/j.cose.2016.06.004

  35. [43]

    J. Huan, W. Wang, and J. Prins. 2003. Efficient mining of frequent subgraphs in the presence of isomorphism. In 3rd IEEE International Conference on Data Mining. https://doi.org/10.1109/icdm.2003.1250974

  36. [44]

    Jiang, F

    C. Jiang, F. Coenen, and M. Zito. 2012. A survey of frequent subgraph mining algorithms. The Knowledge Engineering Review 28, 1 (2012), 75–105. https://doi.org/10.1017/s0269888912000331

  37. [45]

    Jippo, T

    H. Jippo, T. Matsuo, R. Kikuchi, D. Fukuda, A. Matsuura, and M. Ohfuchi. 2019. Graph Classification of Molecules Using Force Field Atom and Bond Types. Molecular Informatics 39, 1-2 (2019), 1800155. https://doi.org/10.1002/minf.201800155

  38. [46]

    Jüttner and P

    A. Jüttner and P. Madarasi. 2018. VF2++ An improved subgraph isomorphism algorithm. Discrete Applied Mathematics 242 (2018), 69–81. https://doi.org/10.1016/j.dam.2018.02.018

  39. [47]

    H. J. Kang and D. Lo. 2022. Active Learning of Discriminative Subgraph Patterns for API Misuse Detection. IEEE Transactions on Software Engineering 48, 8 (2022), 2761–2783. https://doi.org/10.1109/tse.2021.3069978

  40. [48]

    Karbalaie, A

    F. Karbalaie, A. Sami, and M. Ahmadi. 2012. Semantic malware detection by deploying graph mining. International Journal of Computer Science Issues 9, 1 (2012), 373. https://www.researchgate.net/publication/257351721

  41. [49]

    Kashima, K

    H. Kashima, K. Tsuda, and A. H. Inokuchi. 2003. Marginalized kernels between labeled graphs. In 20th international conference on machine learning . 321–328. https://cdn.aaai.org/ICML/2003/ICML03-044.pdf

  42. [50]

    M. G. Kendall. 1938. A new measure of rank correlation. Biometrika 30, 1-2 (1938), 81–93. https://doi.org/10.1093/biomet/30.1-2.81

  43. [51]

    Kikaj, G

    A. Kikaj, G. Marra, and L. De Raedt. 2024. Subgraph Mining for Graph Neural Networks. In International Symposium on Intelligent Data Analysis (Lecture Notes in Computer Science, Vol. 14641) . Springer, 141–152. https://doi.org/10.1007/978-3-031-58547-0_12

  44. [52]

    W. Klösgen. 1996. Explora: a multipattern and multistrategy discovery assistant . American Association for Artificial Intelligence, 249–271. https://dl.acm.org/doi/10.5555/257938.257965

  45. [54]

    N. M. Kriege, P. L. Giscard, and R. Wilson. 2016. On Valid Optimal Assignment Kernels and Applications to Graph Classification. In 30th International Conference on Neural Information Processing Systems . 1623–1631. https://proceedings.neurips.cc/paper_files/paper/2016/hash/ 0e...

  46. [55]

    N. M. Kriege, F. D. Johansson, and C. Morris. 2020. A survey on graph kernels. Applied Network Science 5 (2020), 6. https://doi.org/10.1007/s41109- 019-0195-3

  47. [56]

    Kumar, C

    I. Kumar, C. Scheidegger, S. Venkatasubramanian, and S. Friedler. 2021. Shapley Residuals: Quantifying the limits of the Shapley value for explanations. In Advances in Neural Information Processing Systems , Vol. 34. 26598–26608. https://proceedings.neurips.cc/paper_files/pape...

  48. [57]

    Lavrač, P

    N. Lavrač, P. Flach, and B. Zupan. 1999. Rule Evaluation Measures: A Unifying View. In International Conference on Inductive Logic Programming (Lecture Notes in Computer Science, Vol. 1634) . Springer, 174–185. https://doi.org/10.1007/3-540-48751-4_17

  49. [58]

    Lavrač, B

    N. Lavrač, B. Kavšek, P. Flach, and L. Todorovski. 2004. Subgroup Discovery with CN2-SD. Journal of Machine Learning Research 5 (2004), 153–188. https://www.jmlr.org/papers/v5/lavrac04a.html

  50. [59]

    Lenca, P

    P. Lenca, P. Vaillant, B.and Meyer, and S. Lallich. 2007. Association Rule Interestingness Measures: Experimental and Theoretical Studies. In Quality Measures in Data Mining . Studies in Computational Intelligence, Vol. 43. Springer, 51–76. https://doi.org/10.1007/978-3-540-44918-8_3

  51. [60]

    J. Li, G. Dong, and K. Ramamohanarao. 2000. Making Use of the Most Expressive Jumping Emerging Patterns for Classification. In 4th Pacific-Asia Conference on Knowledge Discovery and Data Mining (Lecture Notes in Computer Science, Vol. 1805) . Springer, 220–232. https://doi.org...

  52. [61]

    P. Li, Y. Xie, X. Xu, J. Zhou, and Q. Xuan. 2022. Phishing Fraud Detection on Ethereum Using Graph Neural Network. In 4th International Conference on Blockchain and Trustworthy Systems (Communications in Computer and Information Science, Vol. 1679) . Springer, 362–375. https: ...

  53. [62]

    Loyola-González, M

    O. Loyola-González, M. A. Medina-Pérez, and K. R. Choo. 2020. A Review of Supervised Classification based on Contrast Patterns: Applications, Trends, and Challenges. Journal of Grid Computing 18, 4 (2020), 797–845. https://doi.org/10.1007/s10723-020-09526-y

  54. [63]

    Loyola-González, M

    O. Loyola-González, M. García-Borroto, J. F. Martínez-Trinidad, and J. A. Carrasco-Ochoa. 2014. An empirical comparison among quality measures for pattern based classifiers. Intelligent Data Analysis 18, 6S (2014), S5–S17. https://doi.org/10.3233/ida-140705

  55. [64]

    S. M. Lundberg, G. Erion, H. Chen, A. DeGrave, J. M. Prutkin, B. Nair, R. Katz, J. Himmelfarb, N. Bansal, and S. Lee. 2020. From local explanations to global understanding with explainable AI for trees. Nature Machine Intelligence 2 (2020), 56–67. https://doi.org/10.1038/s4225...

  56. [65]

    S. M. Lundberg and S. Lee. 2017. A Unified Approach to Interpreting Model Predictions. In 31st International Conference on Neural Information Processing Systems. 4768–4777. https://doi.org/10.5555/3295222.3295230

  57. [66]

    Mathonat, D; Nurbakova, J

    R. Mathonat, D; Nurbakova, J. F. Boulicaut, and M. Kaytoue. 2020. Anytime mining of sequential discriminative patterns in labeled sequences. Knowledge and Information Systems 63, 2 (Nov. 2020), 439–476. https://doi.org/10.1007/s10115-020-01523-7

  58. [67]

    Mélançon

    G. Mélançon. 2006. Just how dense are dense graphs in the real world?: a methodological note. In A VI Workshop on Beyond time and errors: novel evaluation methods for information visualization . 1–7. https://doi.org/10.1145/1168149.1168167

  59. [68]

    D. Q. Nguyen, T. D. Nguyen, and D. Phung. 2022. Universal Graph Transformer Self-Attention Networks. InCompanion Proceedings of the Web Conference. 193–196. https://doi.org/10.1145/3487553.3524258

  60. [69]

    Orsini, P

    F. Orsini, P. Frasconi, and L. De Raedt. 2015. Graph invariant kernels. In 24th International Conference on Artificial Intelligence . 3756–3762. https://doi.org/10.5555/2832747.2832773

  61. [70]

    K. Pearson. 1896. Mathematical Contributions to the Theory of Evolution. III. Regression, Heredity, and Panmixia. Philosophical Transactions of the Royal Society A 187 (1896), 253–318. https://doi.org/10.1098/rsta.1896.0007

  62. [71]

    Piatetsky-Shapiro

    G. Piatetsky-Shapiro. 1991. Discovery, analysis and presentation of strong rules. In Knowledge Discovery in Databases . AAAI Press, Chapter 13, 229–248. https://mitpress.mit.edu/9780262660709/knowledge-discovery-in-databases/

  63. [72]

    Piatetsky-Shapiro and S

    G. Piatetsky-Shapiro and S. Steingold. 2000. Measuring lift quality in database marketing. ACM SIGKDD Explorations Newsletter 2, 2 (2000), 76–80. https://doi.org/10.1145/380995.381018

  64. [73]

    Potin, R

    L. Potin, R. Figueiredo, V. Labatut, and C. Largeron. 2023. Pattern Mining for Anomaly Detection in Graphs: Application to Fraud in Public Procurement. In European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (Lecture Notes in ...

  65. [74]

    Sokal R. R. and C. D. Michener. 1958. A statistical method for evaluating systematic relationships. University of Kansas science bulletin 38 (1958), 1409–1438. http://www.citeulike.org/user/BioNica/article/5845721

  66. [75]

    Ramamohanarao and H

    K. Ramamohanarao and H. Fan. 2007. Patterns Based Classifiers. World Wide Web 10, 1 (2007), 71–83. https://doi.org/10.1007/s11280-006-0012-7

  67. [76]

    Rieck, C

    B. Rieck, C. Bock, and K. Borgwardt. 2019. A Persistent Weisfeiler-Lehman Procedure for Graph Classification. In 36th International Conference on Machine Learning (Proceedings of Machine Learning Research, Vol. 97) . 5448–5458. https://proceedings.mlr.press/v97/rieck19a.html

  68. [77]

    Riesen, K.and Bunke

    H. Riesen, K.and Bunke. 2008. IAM Graph Database Repository for Graph Based Pattern Recognition and Machine Learning. In Joint IAPR International Workshops on Statistical Techniques in Pattern Recognition and Structural and Syntactic Pattern Recognition (Lecture Notes in Compu...

  69. [78]

    Rousseau, E

    F. Rousseau, E. Kiagias, and M. Vazirgiannis. 2015. Text Categorization as a Graph Classification Problem. In 53rd Annual Meeting of the Association for Computational Linguistics / 7th International Joint Conference on Natural Language Processing . 1702–1712. https://doi.org/1...

  70. [79]

    Ahmet Erdem Sarıyüce and Ali Pinar. 2018. Peeling Bipartite Networks for Dense Subgraph Discovery. In 11 ACM International Conference on Web Search and Data Mining . 504–512. https://doi.org/10.1145/3159652.3159678

  71. [80]

    M Sebag and M Schoenauer. 1988. Generation of rules with certainty and confidence factors from incomplete and incoherent learning bases. In European Knowledge Acquisition Workshop, Vol. 88. 28. https://publica.fraunhofer.de/entities/event/b65b8403-dcf9-4edd-9ccd-2b066d3b6608/details

  72. [81]

    C. E. Shannon. 1948. A Mathematical Theory of Communication. Bell System Technical Journal 27, 3 (1948), 379–423. https://doi.org/10.1002/j.1538- 7305.1948.tb01338.x

  73. [82]

    L. S. Shapley. 1953. A Value for n-Person Games. In Contributions to the Theory of Games . Annals of Mathematics Studies, Vol. 28. Princeton University Press, Chapter 17, 307–318. https://doi.org/10.1515/9781400881970-018 Manuscript submitted to ACM Pattern-Based Graph Classif...

  74. [83]

    P. N. Tan, V. Kumar, and J. Srivastava. 2004. Selecting the right objective measure for association analysis.Information Systems 29, 4 (2004), 293–313. https://doi.org/10.1016/s0306-4379(03)00072-3

  75. [84]

    Theng and K

    D. Theng and K. K. Bhoyar. 2023. Feature selection techniques for machine learning: A survey of more than two decades of research. Knowledge and Information Systems 66, 3 (2023), 1575–1637. https://doi.org/10.1007/s10115-023-02010-5

  76. [85]

    Thoma, H

    M. Thoma, H. Cheng, A. Gretton, J. Han, H. Kriegel, and A. Smola. 2009. Near-optimal supervised feature selection among frequent subgraphs. In SIAM International Conference on Data Mining . 1076–1087. https://doi.org/10.1137/1.9781611972795.92

  77. [86]

    Thoma, H

    M. Thoma, H. Cheng, A. Gretton, J. Han, H. P. Kriegel, A. Smola, L. Song, P. S. Yu, S. Yan, and K. M. Borgwardt. 2010. Discriminative frequent subgraph mining with optimality guarantees. Statistical Analysis and Data Mining 3, 5 (2010), 302–318. https://doi.org/10.1002/sam.10084

  78. [87]

    R. M. H. Ting and J. Bailey. 2006. Mining Minimal Contrast Subgraph Patterns. In SIAM International Conference on Data Mining . 639–643. https://doi.org/10.1137/1.9781611972764.76

  79. [88]

    Toivonen, A

    H. Toivonen, A. Srinivasan, R. D. King, S. Kramer, and C. Helma. 2003. Statistical evaluation of the Predictive Toxicology Challenge 2000-2001. Bioinformatics 19, 10 (2003), 1183–1193. https://doi.org/10.1093/bioinformatics/btg130

  80. [89]

    Tsuda and H

    K. Tsuda and H. Saigo. 2010. Graph Classification. In Managing and Mining Graph Data . Advances in Database Systems, Vol. 40. Springer, 337–363. https://doi.org/10.1007/978-1-4419-6045-0_11

  81. [90]

    C. J. van Rijsbergen. 1979. Information Retrieval. Butterworths. https://doi.org/10.1002/asi.4630300621

  82. [91]

    Ventura and J

    S. Ventura and J. M. Luna. 2016. Quality Measures in Pattern Mining . Springer, 27–44. https://doi.org/10.1007/978-3-319-33858-3_2

  83. [92]

    Veyrin-Forrer, A

    L. Veyrin-Forrer, A. Kamal, S. Duffner, M. Plantevit, and C. Robardet. 2022. In pursuit of the hidden features of GNN’s internal representations. Data & Knowledge Engineering 142 (2022), 102097. https://doi.org/10.1016/j.datak.2022.102097

  84. [93]

    Wale and G

    N. Wale and G. Karypis. 2006. Comparison of Descriptor Spaces for Chemical Compound Retrieval and Classification. In6th International Conference on Data Mining. 678–689. https://doi.org/10.1109/icdm.2006.39

  85. [94]

    G. I. Webb and S. Zhang. 2005. K-Optimal Rule Discovery.Data Mining and Knowledge Discovery 10, 1 (2005), 39–79. https://doi.org/10.1007/s10618- 005-0255-4

  86. [95]

    Webber, A

    W. Webber, A. Moffat, and J. Zobel. 2010. A similarity measure for indefinite rankings. ACM Transactions on Information Systems 28, 4 (2010), 1–38. https://doi.org/10.1145/1852102.1852106

  87. [96]

    B. Wu, Y. Liu, B. Lang, and L. Huang. 2018. DGCNN: Disordered graph convolutional neural network based on the Gaussian mixture model. Neurocomputing 321 (2018), 346–356. https://doi.org/10.1016/j.neucom.2018.09.008

  88. [97]

    D. Wu, Q. Wang, and D. L. Olson. 2023. Industry classification based on supply chain network information using Graph Neural Networks. Applied Soft Computing 132 (2023), 109849. https://doi.org/10.1016/j.asoc.2022.109849

  89. [98]

    J. Wu, S. Pan, X. Zhu, and Z. Cai. 2015. Boosting for Multi-Graph Classification. IEEE Transactions on Cybernetics 45, 3 (2015), 416–429. https://doi.org/10.1109/tcyb.2014.2327111

  90. [99]

    Z. Wu, S. Pan, F. Chen, G. Long, C. Zhang, and P. S. Yu. 2021. A Comprehensive Survey on Graph Neural Networks. IEEE Transactions on Neural Networks and Learning Systems 32, 1 (2021), 4–24. https://doi.org/10.1109/tnnls.2020.2978386

  91. [100]

    Yan and J

    X. Yan and J. Han. 2002. gSpan: graph-based substructure pattern mining. In IEEE International Conference on Data Mining . 721–724. https: //doi.org/10.1109/ICDM.2002.1184038

  92. [101]

    Yanardag and S.V.N

    P. Yanardag and S.V.N. Vishwanathan. 2015. Deep Graph Kernels. In 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 1365–1374. https://doi.org/10.1145/2783258.2783417

  93. [102]

    Yang and T

    Y. Yang and T. O. Pedersen. 1997. A Comparative Study on Feature Selection in Text Categorization. In14th International Conference on Machine Learning. 412–420. https://doi.org/10.5555/645526.657137

  94. [103]

    Yin and J

    X. Yin and J. Han. 2003. CPAR: Classification based on Predictive Association Rules. In SIAM International Conference on Data Mining . 331–335. https://doi.org/10.1137/1.9781611972733.40

  95. [104]

    Zhang and S

    C. Zhang and S. Zhang. 2002. Association Rule Mining: Models and Algorithms . Lecture Notes in Computer Science, Vol. 2307. Springer. https: //doi.org/10.1007/3-540-46027-6

  96. [105]

    Zhang and X

    S. Zhang and X. Wu. 2011. Fundamentals of association rules in data mining and knowledge discovery. WIREs Data Mining and Knowledge Discovery 1, 2 (2011), 97–116. https://doi.org/10.1002/widm.10

  97. [106]

    J. Zhou, G. Cui, S. Hu, Z. Zhang, C. Yang, Z. Liu, L. Wang, C. Li, and M. Sun. 2020. Graph neural networks: A review of methods and applications. AI Open 1 (2020), 57–81. https://doi.org/10.1016/j.aiopen.2021.01.001 A ADDITIONAL INFORMATION ABOUT QUALITY MEASURES This appendix...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.