Pith. sign in

REVIEW 5 major objections 5 minor 70 references

DeepSN: A Sheaf Neural Framework for Influence Maximization

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read DeepSN claims a sheaf reaction-diffusion GNN with Louvain-based seed allocation matches or beats state-of-the-art influence maximization, with large SIS gains.

desk verdict New sheaf diffusion-reaction architecture for influence maximization with strong SIS numbers, but the headline claim is internally contradicted and the theoretical appendix needs correction. read the letter →

arxiv 2412.12416 v1 pith:S6KTLQVN submitted 2024-12-16 cs.LG cs.AIcs.SI

Asela Hevapathige , Qing Wang , Ahad N. Zehmakan This is my paper · ORCID
classification cs.LGcs.AIcs.SI
keywords influencemaximizationsheafneuralnetworksgraphreaction-diffusionseedselectionSISmodelLouvaincommunitydetectiondiffusionmodels
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 introduces DeepSN, a graph neural network that treats influence propagation as a sheaf reaction-diffusion process and learns to pick seed sets that maximize spread. The central claim is that combining a sheaf Laplacian with learnable vertex-confidence weights and two reaction operators lets the model capture both progressive (IC, LT) and non-progressive (SIS) diffusion dynamics, where standard GNNs oversmooth and fail. For seed selection, DeepSN builds a weighted graph from the learned sheaf coefficients, partitions it with Louvain, and allocates seeds proportionally to subgraph sizes, shrinking the combinatorial search space. In experiments on graphs up to roughly 280,000 vertices, DeepSN matches or beats state-of-the-art methods on IC and LT and reports substantial gains on SIS. If the claims hold, learning-based influence maximization can be made both diffusion-agnostic and scalable to real networks.

What carries the argument

The load-bearing objects are the nonlinear sheaf Laplacian with learnable trust coefficients and the two reaction operators that turn pure diffusion into reaction-diffusion. The Laplacian measures, per edge, the disagreement between two vertices' public opinions after translation by the restriction maps; the reaction terms add bounded, non-monotonic updates and remove the forced equality of transformed features at the fixed point ($F_{v\unlhd e}x_v = F_{u\unlhd e}x_u$), which is what gives the model separability and protects against oversmoothing. On the optimization side, the key mechanism is building the weighted graph $G_w$ from the learned sheaf coefficients, applying Louvain to obtain subgraphs that are assumed to have low cross-subgraph influence overlap, and allocating seed budget proportionally to subgraph size.

What would settle it

Take a real network such as Cora-ML under the IC model, compute the Louvain partition of the learned weighted graph $G_w$, and measure the expected number of vertices reachable from seeds in two different subgraphs that are also reachable from one another. If that cross-subgraph overlap is large, or if replacing the per-subgraph constraint $|S_i| \leq (k/n)|V_i|$ with global seed selection gives equal or larger total influence, the central seed-allocation claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that influence maximization can be solved end-to-end by a sheaf reaction-diffusion GNN coupled to a subgraph-based seed optimizer. Influence propagation is modeled as a sheaf diffusion process: each vertex carries a private opinion vector, each edge carries a public-opinion space, and restriction maps translate private opinions to edge level; a learnable coefficient $\psi_{vu}$ in the nonlinear sheaf Laplacian encodes how much vertex $v$ trusts neighbor $u$. Two reaction operators, pointwise dynamics for vertex-intrinsic state changes and coupled dynamics for the combined effect of activated and susceptible neighbors, make the evolution non-monotonic and, the paper argues, capture non-progressive models such as SIS. After training on ground-truth activation probabilities, the learned sheaf coefficients define a weighted graph $G_w$, which is partitioned by Louvain into subgraphs; a small MLP then selects seeds within each subgraph subject to $|S_i| \leq (k/n)|V_i|$, and total influence is estimated by the trained sheaf GNN. The reported outcome is that DeepSN and its sparsified variant DeepSN_SP match or outperform IMM, OPIM, SubSIM, IMINFECTOR, PIANO, ToupleGDD, and DeepIM across IC, LT, and SIS, with the largest relative gains on SIS.

Load-bearing premise

The argument assumes that partitioning the sheaf-coefficient weighted graph with Louvain actually minimizes the overlap of influence between vertices in different subgraphs, so that allocating seeds proportionally to subgraph size does not sacrifice spread.

Editorial extensions

If this is right

  • If the sheaf reaction-diffusion model genuinely captures non-progressive dynamics, a single trained DeepSN can be applied to diffusion models with recovery or oscillation without task-specific re-engineering.
  • The Louvain-based reduction turns the exponential seed search into per-community selection, which is what lets the method scale to graphs of roughly 280,000 vertices.
  • The relaxation of the sheaf fixed-point condition implies the architecture can keep neighboring vertex features distinct, a property the paper links to stronger estimation of long-range influence.
  • Because the diffusion model is learned from ground-truth activation data rather than supplied, the same framework can switch among IC, LT, and SIS settings without changing the pipeline.

Reading between the lines

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

  • If the Louvain overlap assumption is weak on real networks, proportional seed allocation could underperform global selection; a natural extension is to learn per-subgraph budgets instead of fixing $|S_i| \leq (k/n)|V_i|$.
  • The reaction operators resemble epidemic and opinion-dynamics terms, so the architecture may transfer to multi-stage or multi-state models such as SEIR or threshold models with more than two states, which the paper lists as future work.
  • DeepSN's gains on SIS suggest that other non-progressive phenomena, like recurring misinformation or churn in user activity, could be treated with the same estimator.
  • Since baseline numbers are taken from a single prior study, an independent replication or an evaluation against freshly run traditional algorithms would clarify whether the improvement comes from the sheaf estimator, the Louvain seed allocation, or the experimental setup.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper proposes DeepSN, a two-phase framework for influence maximization. In the first phase, a sheaf reaction-diffusion GNN is trained to estimate vertex activation probabilities under IC, LT, and SIS diffusion models. In the second phase, seed sets are selected by applying Louvain community detection to a graph weighted by learned sheaf coefficients and then allocating seeds across subgraphs via an MLP. The authors claim that DeepSN and a sparsified variant DeepSNSP outperform or match state-of-the-art baselines, with a substantial improvement under SIS, and provide stability and separability lemmas and propositions in support of the architecture.

Significance. If the claims held, the paper would make a useful contribution to learning-based influence maximization: it targets non-progressive SIS dynamics, scales to graphs of roughly 280,000 vertices, and offers a computationally cheaper variant. Strengths include the breadth of experiments across six datasets and three diffusion models, ablations of the reaction terms and seed-allocation components, and a complexity analysis showing only a constant overhead for the feature dimensions used. The manuscript is also candid about its current limitation of supporting only two-state diffusion models. However, the theoretical support contains load-bearing errors, the main SIS claim is contradicted by the paper's own Table 1, and the seed-selection principle rests on an unvalidated assumption. These issues need to be resolved before the empirical results can be taken at face value.

major comments (5)
  1. [Experiments (Exp-1), Table 1] The statement that 'DeepSN achieves a substantial improvement over all baseline methods for the SIS diffusion model' is contradicted by the numbers reported in Table 1. In the Network Science SIS block at a 1% seed budget, DeepSN reports 2.0% total influence versus DeepIM's 2.7%. In the Digg SIS block at a 20% seed budget, DeepSN reports 33.2% and DeepSNSP reports 33.1% versus DeepIM's 36.3%. Thus the unqualified 'all baselines' claim is not supported by the paper's own data. Please revise the claim, report uncertainty or error bars, and specify the exact SIS evaluation protocol.
  2. [Appendix, proof of Lemma 3] The proof of Lemma 3 derives ||ΔF X†||2 ≤ C and then uses the largest eigenvalue λmax of ΔF to conclude ||X†||2 ≤ C/(|α|λmax). This is a direction error: the operator norm bound ||ΔF X†|| ≤ λmax ||X†|| cannot be inverted to upper-bound ||X†||. The correct argument would use the smallest eigenvalue, giving ||X†|| ≤ ||ΔF X†||/λmin when ΔF is symmetric positive definite. As written, the proof of boundedness of the fixed point is invalid.
  3. [Appendix, proof of Proposition 3] The proof of Proposition 3 divides the fixed-point equation by X† and treats αΔF as a scalar to impose the condition 0 < αΔF < a/b + c/d. Here X† is a vector (or stacked matrix), ΔF is an operator, and the reaction terms are not scalar functions of X†. The monotonicity argument for f(X†) does not establish existence or uniqueness of a fixed point for the resulting vector-valued nonlinear equation. Consequently Proposition 3 and the separability conclusion of Remark 2 are unproven.
  4. [Optimizing Seed Selection] The paper asserts that applying Louvain to the sheaf-weighted graph Gw 'minimizes the overlap of influence between vertices in different subgraphs.' Louvain maximizes modularity, which is a structural objective; no argument or experiment is provided connecting modularity communities to low influence overlap under IC, LT, or SIS. In addition, the constraint |Si| ≤ (k/n)|Vi| forces seed counts proportional to community size, which is not obviously optimal when communities differ in influence potential. This is a load-bearing premise of the seed-selection phase and should be validated or substantially weakened.
  5. [Appendix, Experimental Setups] The evaluation protocol says each diffusion model is simulated 'until the diffusion process halts, or converges to a steady-state.' For a non-progressive SIS model this stopping rule is ambiguous: the process can persist indefinitely, and the infection and recovery rates are not specified in the paper. Since the SIS baseline numbers are taken from Ling et al. (2023) rather than recomputed, the comparability of DeepSN's SIS results depends on exactly matching an unspecified protocol. Please provide the SIS parameters and the precise stopping criterion, and ideally recompute baselines under the same protocol.
minor comments (5)
  1. [Table 1 and Table 4] The tables are difficult to read because values run together without visual separators; for example, in Table 1 the IMINfECTOR IC row contains '50.92.1' where two distinct numbers are merged, and similar merged entries appear for DeepSN in the Power Grid IC row. Please regenerate the tables with clear cell boundaries.
  2. [Lemma 2 proof] The proof of Lemma 2 compares Φ1_v,i X / (κ + |X|) with Φ1_v,i as though Φ1_v,i is nonnegative; if these coefficients can be negative, the inequality should use |Φ1_v,i| or explicit sign conditions.
  3. [Proposition 2 proof] The proof relies on Proposition 9 of Bodnar et al. (2022) and on 'symmetric sheaves,' but the latter notion is not defined in this manuscript; please add a definition or a precise citation.
  4. [Equation (11), training of Tφ] The training objective for the seed-selection network Tφ involves the influence σ of the selected seed set, but the paper does not explain how gradients flow through the discrete selection operation or how Tφ is trained in practice; a short training description would clarify the method.
  5. [References and formatting] The reference to Caralt et al. is incomplete (no year or venue), and the Limitations section is a single sentence; please expand both.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the framework is trained on ground-truth activation data and evaluated by independent Monte Carlo simulation; no load-bearing step reduces to its own inputs.

full rationale

The core derivation is self-contained rather than circular. The GNN is trained to minimize MSE against ground-truth activation probabilities (Eq. 10), the seed selector Tφ is trained to maximize the GNN's predicted influence (Eq. 11), and the reported Total Influence values are obtained by independently simulating each diffusion model for 100 repetitions until the process halts or reaches steady state, as described in the Experimental Setups. The final benchmark is thus not the training objective, and no fitted parameter is relabeled as a prediction. The fixed-point and separability claims (Propositions 1-3) are proved from the reaction-diffusion equations with the help of external theorems (Hansen and Ghrist 2021; Bodnar et al. 2022); they are not inherited from the authors' own prior work, and there is no self-citation chain that forces the architecture or the results. Two issues are worth flagging, but they are correctness questions rather than circularity. First, Exp-1 states that DeepSN 'achieves a substantial improvement over all baseline methods for the SIS diffusion model,' yet Table 1 shows that on Digg at 20% budget DeepSN reaches 33.2% total influence while DeepIM reaches 36.3%, so the paper's own data contradict the unqualified claim. Second, the optimization section asserts that Louvain on the sheaf-coefficient weighted graph 'minimizes the overlap of influence between vertices in different subgraphs,' but this is an unverified heuristic, not a theorem or an evaluation, and no experiment isolates it. Additionally, the SIS termination criterion is ambiguous for a non-progressive model and baselines are copied from Ling et al. (2023) rather than recomputed, which weakens comparability without making the derivation circular. The proof of Proposition 3 also contains apparent notational gaps, such as dividing by the vector X† and comparing the matrix α∆F to scalars; these are proof-correctness concerns, not circularity.

Assumptions & free parameters 8 free parameters · 7 assumptions · 0 invented entities

The central claim rests on learned parameters and modeling assumptions rather than on derived physical or mathematical constants. The most fragile assumptions are the opinion-diffusion analogy for influence and the Louvain overlap-minimization premise. The ϵ-shift assumption is internally inconsistent with the harmonic convergence result. No new physical entities are introduced.

free parameters (8)
  • learnable sheaf coefficients ψ_vu = learned; final values not reported
    Control trust/edge weighting in Eq. (2) and define the weighted graph Gw used for Louvain partitioning.
  • reaction operator coefficients Φ1_v, Φ2_v, κ1_v, κ2_v = learned; final values not reported
    Define pointwise and coupled reaction operators in Eq. (7); boundedness proofs assume κ > 0.
  • GNN weight matrices W1^t, W2^t = learned; not reported
    Layer-wise feature transformations in Eq. (8).
  • diffusion/reaction weights α, β, γ = not reported
    Balance diffusion, pointwise, and coupled contributions in Eq. (7); treated as hand-chosen coefficients.
  • positive-definite shift epsilon = not reported
    Added to the Laplacian in Eq. (3) to force positive definiteness; its use conflicts with harmonic convergence in Proposition 1.
  • Louvain resolution parameter = searched in {0.1, 1, 2}; final value not reported
    Controls subgraph granularity and therefore the number of seeds allocated to each community.
  • DeepSNSP sparsification threshold = 0.5
    Converts continuous sheaf coefficients to binary values for the computationally efficient variant.
  • MLP hidden units, layers, dropout, learning rate, batch size = searched; final values not reported
    Training hyperparameters whose chosen values directly affect the reported performance numbers.
assumptions (7)
  • standard math Standard cellular sheaf and sheaf Laplacian definitions (Definitions 2 and 3) are used as the modeling foundation.
    The paper builds on established sheaf theory from Tennison and Bredon, and on sheaf diffusion GNNs from Bodnar et al.
  • domain assumption Influence propagation can be represented as opinion diffusion on a cellular sheaf, where vertex sheaves are private opinions and edge sheaves are public opinions.
    Introduced in the Learning to Estimate Influence section; no empirical or theoretical justification connects this analogy to IC, LT, or SIS dynamics.
  • domain assumption The reaction-diffusion PDE in Eq. (7) with pointwise and coupled reaction operators accurately models progressive and non-progressive influence diffusion.
    The architecture is designed to fit these operators, but the paper does not prove they can represent IC, LT, or SIS dynamics.
  • ad hoc to paper Adding epsilon I makes the sheaf Laplacian positive definite, and positive definiteness is the relevant condition for convergence of the diffusion process.
    Eq. (3) introduces this modification, but Proposition 1 still relies on convergence to H0, the kernel of the unmodified Laplacian, creating an internal inconsistency.
  • ad hoc to paper Louvain communities on the sheaf-weighted graph Gw minimize overlap of influence between vertices in different subgraphs.
    Stated in Optimizing Seed Selection without proof or targeted evaluation; this is the load-bearing assumption for proportional seed allocation.
  • domain assumption Ground-truth activation probabilities Y are available for supervised training.
    The MSE loss in Eq. (10) requires ground-truth activation probabilities for all vertices, which the experiments generate from simulations.
  • standard math Theorem 2.2 of Hansen and Ghrist (2021) on convergence of sheaf diffusion to H0(G;F) applies unchanged to the setup in Proposition 1.
    The proof of Proposition 1 invokes this theorem without checking that the hypotheses hold after the ϵ-shift modification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DeepSN: A Sheaf Neural Framework for Influence Maximization." pith.science (2026). https://pith.science/paper/S6KTLQVN

@misc{pith2026241212416,
  author       = {Pith},
  title        = {Pith review of: DeepSN: A Sheaf Neural Framework for Influence Maximization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6KTLQVN}},
  note         = {Machine review of arXiv:2412.12416}
}
read the original abstract

Influence maximization is key topic in data mining, with broad applications in social network analysis and viral marketing. In recent years, researchers have increasingly turned to machine learning techniques to address this problem. They have developed methods to learn the underlying diffusion processes in a data-driven manner, which enhances the generalizability of the solution, and have designed optimization objectives to identify the optimal seed set. Nonetheless, two fundamental gaps remain unsolved: (1) Graph Neural Networks (GNNs) are increasingly used to learn diffusion models, but in their traditional form, they often fail to capture the complex dynamics of influence diffusion, (2) Designing optimization objectives is challenging due to combinatorial explosion when solving this problem. To address these challenges, we propose a novel framework, DeepSN. Our framework employs sheaf neural diffusion to learn diverse influence patterns in a data-driven, end-to-end manner, providing enhanced separability in capturing diffusion characteristics. We also propose an optimization technique that accounts for overlapping influence between vertices, which helps to reduce the search space and identify the optimal seed set effectively and efficiently. Finally, we conduct extensive experiments on both synthetic and real-world datasets to demonstrate the effectiveness of our framework.

Figures

Figures reproduced from arXiv: 2412.12416 by the authors.

Figure 1
Figure 1. The DeepSN framework consists of two phases: (a) learning to estimate influence with sheaf GNN; b) optimizing [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Performance of DeepSN for influence estimation in terms of MAE (Mean Absolute Error), compared to baseline [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Impact of layer depth and feature dimension on [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Performance of DeepSN for influence estimation under IC model. [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

70 extracted references · 63 canonical work pages

  1. [1]

    Banerjee, S.; Jenamani, M.; and Pratihar, D. K. 2020. A survey on influence maximization in a social network. Knowledge and Information Systems, 62

    work page 2020

  2. [2]

    Bao, F.; Zhao, M.; Hao, Z.; Li, P.; Li, C.; and Zhu, J. 2022. Equivariant energy-guided sde for inverse molecular design. In The eleventh international conference on learning representations

    work page 2022

  3. [3]

    S.; Bronstein, M.; Veli c kovi \'c , P.; and Li \`o , P

    Barbero, F.; Bodnar, C.; de Oc \'a riz Borde, H. S.; Bronstein, M.; Veli c kovi \'c , P.; and Li \`o , P. 2022. Sheaf neural networks with connection laplacians. In Topological, Algebraic and Geometric Learning Workshops 2022. PMLR

    work page 2022

  4. [4]

    Bellman, R. 1947. On the boundedness of solutions of nonlinear differential and difference equations. Transactions of the American Mathematical Society, 62(3)

    work page 1947

  5. [5]

    D.; Guillaume, J.-L.; Lambiotte, R.; and Lefebvre, E

    Blondel, V. D.; Guillaume, J.-L.; Lambiotte, R.; and Lefebvre, E. 2008. Fast unfolding of communities in large networks. Journal of statistical mechanics: theory and experiment, 2008(10)

    work page 2008

  6. [6]

    Bodnar, C.; Di Giovanni, F.; Chamberlain, B.; Lio, P.; and Bronstein, M. 2022. Neural sheaf diffusion: A topological perspective on heterophily and oversmoothing in gnns. Advances in Neural Information Processing Systems, 35

    work page 2022

  7. [7]

    Bredon, G. E. 2012. Sheaf theory, volume 170. Springer Science & Business Media

    work page 2012

  8. [8]

    H.; Bernardez, G.; Duta, I.; Alarcon, E.; and Lio, P

    Caralt, F. H.; Bernardez, G.; Duta, I.; Alarcon, E.; and Lio, P. ???? Complex Diffusion Processes as an Inductive Bias in Sheaf Neural Networks. In ICML 2024 Workshop on Geometry-grounded Representation Learning and Generative Modeling

    work page 2024

Show all 70 references
  1. [9]

    I.; Bronstein, M.; Webb, S.; and Rossi, E

    Chamberlain, B.; Rowbottom, J.; Gorinova, M. I.; Bronstein, M.; Webb, S.; and Rossi, E. 2021. Grand: Graph neural diffusion. In International conference on machine learning. PMLR

    2021

  2. [10]

    Chen, D.; Lin, Y.; Li, W.; Li, P.; Zhou, J.; and Sun, X. 2020. Measuring and relieving the over-smoothing problem for graph neural networks from the topological view. In Proceedings of the AAAI conference on artificial intelligence, volume 34

    2020

  3. [11]

    Chen, T.; Yan, S.; Guo, J.; and Wu, W. 2023. ToupleGDD: A fine-designed solution of influence maximization by deep reinforcement learning. IEEE Transactions on Computational Social Systems, 11(2)

    2023

  4. [12]

    Chen, W.; Castillo, C.; and Lakshmanan, L. V. 2022. Information and influence propagation in social networks. Springer Nature

    2022

  5. [13]

    Chen, W.; Peng, B.; Schoenebeck, G.; and Tao, B. 2022. Adaptive greedy versus non-adaptive greedy for influence maximization. Journal of Artificial Intelligence Research, 74

    2022

  6. [14]

    Choi, J.; Hong, S.; Park, N.; and Cho, S.-B. 2023. Gread: Graph neural reaction-diffusion networks. In International Conference on Machine Learning. PMLR

    2023

  7. [15]

    Dugu \'e , N.; and Perez, A. 2015. Directed Louvain: maximizing modularity in directed networks. Ph.D. thesis, Universit \'e d'Orl \'e ans

    2015

  8. [16]

    Duta, I.; Cassar \`a , G.; Silvestri, F.; and Li \`o , P. 2024. Sheaf hypergraph networks. Advances in Neural Information Processing Systems, 36

    2024

  9. [17]

    d’Onofrio, A. 2008. A note on the global behaviour of the network-based SIS epidemic model. Nonlinear Analysis: Real World Applications, 9(4)

    2008

  10. [18]

    Eliasof, M.; Haber, E.; and Treister, E. 2024. Feature transportation improves graph neural networks. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38

    2024

  11. [19]

    Erdős, P.; and Rényi, A. 1959. On random graphs I. Publ. math. debrecen, 6(290-297)

    1959

  12. [20]

    Fan, C.; Jiang, Y.; and Mostafavi, . A. 2021. The role of local influential users in spread of situational crisis information. Journal of Computer-Mediated Communication, 26(2)

    2021

  13. [21]

    Gasteiger, J.; Wei enberger, S.; and G \"u nnemann, S. 2019. Diffusion improves graph learning. Advances in neural information processing systems, 32

    2019

  14. [22]

    Goldenberg, J.; Libai, B.; and Muller, E. 2001. Using complex systems analysis to advance marketing theory development: Modeling heterogeneity effects on new product growth through stochastic cellular automata. Academy of Marketing Science Review, 9(3)

    2001

  15. [23]

    Granovetter, M. 1978. Threshold models of collective behavior. American journal of sociology, 83(6)

    1978

  16. [24]

    Guo, Q.; Wang, S.; Wei, Z.; and Chen, M. 2020. Influence maximization revisited: Efficient reverse reachable set generation with bound tightened. In Proceedings of the 2020 ACM SIGMOD international conference on management of data

    2020

  17. [25]

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

    2017

  18. [26]

    Hansen, J.; and Gebhart, T. 2020. Sheaf neural networks. arXiv preprint arXiv:2012.06333

    2020 arXiv

  19. [27]

    Hansen, J.; and Ghrist, R. 2021. Opinion dynamics on discourse sheaves. SIAM Journal on Applied Mathematics, 81(5)

    2021

  20. [28]

    Hosseini-Pozveh, M.; Zamanifar, K.; and Naghsh-Nilchi, A. R. 2017. A community-based approach to identify the most influential nodes in social networks. Journal of Information Science, 43(2)

    2017

  21. [29]

    Kempe, D.; Kleinberg, J.; and Tardos, \'E . 2003. Maximizing the spread of influence through a social network. In Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining

    2003

  22. [30]

    Khoshraftar, S.; and An, A. 2024. A survey on graph representation learning methods. ACM Transactions on Intelligent Systems and Technology, 15(1)

    2024

  23. [31]

    Kingma, D. P. 2015. Adam: A method for stochastic optimization. International Conference on Learning Representations

    2015

  24. [32]

    N.; and Welling, M

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

    2016 arXiv

  25. [33]

    Kumar, S.; Mallik, A.; Khetarpal, A.; and Panda, B. S. 2022. Influence maximization in social networks using graph embedding and graph neural network. Information Sciences, 607

    2022

  26. [34]

    Lancichinetti, A.; and Fortunato, S. 2009. Community detection algorithms: a comparative analysis. Physical Review E—Statistical, Nonlinear, and Soft Matter Physics, 80(5)

    2009

  27. [35]

    B.; Rossi, R

    Lee, J. B.; Rossi, R. A.; Kim, S.; Ahmed, N. K.; and Koh, E. 2019. Attention models in graphs: A survey. ACM Transactions on Knowledge Discovery from Data (TKDD), 13(6)

    2019

  28. [36]

    Lerman, K.; and Galstyan, A. 2008. Analysis of social voting patterns on digg. In Proceedings of the first workshop on Online social networks

    2008

  29. [37]

    Leskovec, J.; Krause, A.; Guestrin, C.; Faloutsos, C.; VanBriesen, J.; and Glance, N. 2007. Cost-effective outbreak detection in networks. In Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining

    2007

  30. [38]

    S.; Rayhan, J

    Li, H.; Xu, M.; Bhowmick, S. S.; Rayhan, J. S.; Sun, C.; and Cui, J. 2022. PIANO: Influence maximization meets deep reinforcement learning. IEEE Transactions on Computational Social Systems, 10(3)

    2022

  31. [39]

    D.; Dinh, T

    Li, X.; Smith, J. D.; Dinh, T. N.; and Thai, M. T. 2019. Tiptop:(almost) exact solutions for influence maximization in billion-scale networks. IEEE/ACM Transactions on Networking, 27(2)

    2019

  32. [40]

    Li, Y.; Fan, J.; Wang, Y.; and Tan, K.-L. 2018. Influence maximization on social graphs: A survey. IEEE Transactions on Knowledge and Data Engineering, 30(10)

    2018

  33. [41]

    Li, Y.; Gao, H.; Gao, Y.; Guo, J.; and Wu, W. 2023. A survey on influence maximization: From an ml-based combinatorial optimization. ACM Transactions on Knowledge Discovery from Data, 17(9)

    2023

  34. [42]

    Li, Y.-M.; Lai, C.-Y.; and Lin, C.-H. 2009. Discovering influential nodes for viral marketing. In 2009 42nd Hawaii International Conference on System Sciences. IEEE

    2009

  35. [43]

    T.; Xue, R.; Song, J.; Qiu, M.; and Zhao, L

    Ling, C.; Jiang, J.; Wang, J.; Thai, M. T.; Xue, R.; Song, J.; Qiu, M.; and Zhao, L. 2023. Deep graph representation learning and optimization for influence maximization. In International Conference on Machine Learning. PMLR

    2023

  36. [44]

    Liu, C.; Fan, W.; Liu, Y.; Li, J.; Li, H.; Liu, H.; Tang, J.; and Li, Q. 2023. Generative diffusion models on graphs: methods and applications. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence

    2023

  37. [45]

    Y.; Bhagat, S.; Lakshmanan, L

    Lou, V. Y.; Bhagat, S.; Lakshmanan, L. V.; and Vaswani, S. 2014. Modeling non-progressive phenomena for influence propagation. In Proceedings of the second ACM conference on Online social networks

    2014

  38. [46]

    C.; and Nandi, A

    Ma, L.; Shao, Z.; Li, X.; Lin, Q.; Li, J.; Leung, V. C.; and Nandi, A. K. 2022. Influence maximization in complex networks by using evolutionary deep reinforcement learning. IEEE Transactions on Emerging Topics in Computational Intelligence, 7(4)

    2022

  39. [47]

    A.; Wilson, P.; Ridler, A.; and Heuer, C

    Marquetoux, N.; Stevenson, M. A.; Wilson, P.; Ridler, A.; and Heuer, C. 2016. Using social network analysis to inform disease control interventions. Preventive Veterinary Medicine, 126

    2016

  40. [48]

    K.; Nigam, K.; Rennie, J.; and Seymore, K

    McCallum, A. K.; Nigam, K.; Rennie, J.; and Seymore, K. 2000. Automating the construction of internet portals with machine learning. Information Retrieval, 3

    2000

  41. [49]

    Nguyen, B.; Sani, L.; Qiu, X.; Li \`o , P.; and Lane, N. D. 2024. Sheaf HyperNetworks for Personalized Federated Learning. arXiv preprint arXiv:2405.20882

    2024 arXiv

  42. [50]

    Niu, C.; Song, Y.; Song, J.; Zhao, S.; Grover, A.; and Ermon, S. 2020. Permutation invariant graph generation via score-based generative modeling. In International Conference on Artificial Intelligence and Statistics. PMLR

    2020

  43. [51]

    D.; and Vazirgianis, M

    Panagopoulos, G.; Malliaros, F. D.; and Vazirgianis, M. 2020. Influence maximization using influence and susceptibility embeddings. In Proceedings of the International AAAI Conference on Web and Social Media, volume 14

    2020

  44. [52]

    D.; and Vazirgiannis, M

    Panagopoulos, G.; Malliaros, F. D.; and Vazirgiannis, M. 2020. Multi-task learning for influence estimation and maximization. IEEE Transactions on Knowledge and Data Engineering, 34(9)

    2020

  45. [53]

    Panagopoulos, G.; Tziortziotis, N.; Vazirgiannis, M.; and Malliaros, F. 2023. Maximizing influence with graph neural networks. In Proceedings of the International Conference on Advances in Social Networks Analysis and Mining

    2023

  46. [54]

    Rossi, R.; and Ahmed, N. 2015. The network data repository with interactive graph analytics and visualization. In Proceedings of the AAAI conference on artificial intelligence, volume 29

    2015

  47. [55]

    Sastry, S. 2013. Nonlinear systems: analysis, stability, and control, volume 10. Springer Science & Business Media

    2013

  48. [56]

    Tang, J.; Tang, X.; Xiao, X.; and Yuan, J. 2018. Online processing algorithms for influence maximization. In Proceedings of the 2018 international conference on management of data

    2018

  49. [57]

    Tang, Y.; Shi, Y.; and Xiao, X. 2015. Influence maximization in near-linear time: A martingale approach. In Proceedings of the 2015 ACM SIGMOD international conference on management of data

    2015

  50. [58]

    Tennison, B. R. 1975. Sheaf theory, volume 21. Cambridge University Press

    1975

  51. [59]

    A.; Waltman, L.; and Van Eck, N

    Traag, V. A.; Waltman, L.; and Van Eck, N. J. 2019. From Louvain to Leiden: guaranteeing well-connected communities. Scientific reports, 9(1)

    2019

  52. [60]

    Turing, A. M. 1990. The chemical basis of morphogenesis. Bulletin of mathematical biology, 52

    1990

  53. [61]

    Vabalas, A.; Gowen, E.; Poliakoff, E.; and Casson, A. J. 2019. Machine learning algorithm validation with a limited sample size. PloS one, 14(11)

    2019

  54. [62]

    Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Li \`o , P.; and Bengio, Y. 2018. Graph Attention Networks. In International Conference on Learning Representations

    2018

  55. [63]

    Wang, C.; Liu, Y.; Gao, X.; and Chen, G. 2021. A reinforcement learning model for influence maximization in social networks. In International Conference on Database Systems for Advanced Applications. Springer

    2021

  56. [64]

    Wang, Y.; Cong, G.; Song, G.; and Xie, K. 2010. Community-based greedy algorithm for mining top-k influential nodes in mobile social networks. In Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining

    2010

  57. [65]

    G.; and Jin, S

    Wang, Y.; Yi, K.; Liu, X.; Wang, Y. G.; and Jin, S. 2022. ACMP: Allen-cahn message passing with attractive and repulsive forces for graph neural networks. In The Eleventh International Conference on Learning Representations

    2022

  58. [66]

    Xia, W.; Li, Y.; Wu, J.; and Li, S. 2021. Deepis: Susceptibility estimation on social networks. In Proceedings of the 14th ACM International Conference on Web Search and Data Mining

    2021

  59. [67]

    Zhang, S.; Tong, H.; Xu, J.; and Maciejewski, R. 2019. Graph convolutional networks: a comprehensive review. Computational Social Networks, 6(1)

    2019

  60. [68]

    Zhao, J.; Dong, Y.; Ding, M.; Kharlamov, E.; and Tang, J. 2021. Adaptive diffusion in graph neural networks. Advances in neural information processing systems, 34

    2021

  61. [69]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  62. [70]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.