Pith. sign in

REVIEW 5 major objections 9 minor 74 references

REM: A Scalable Reinforced Multi-Expert Framework for Multiplex Influence Maximization

T0 review · 5 major / 9 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read REM claims multiplex influence maximization reduces to optimizing a continuous latent seed vector, beating six baselines on five real networks.

desk verdict Plausible system, invalid theory, and tables that contradict the abstract: this MIM paper needs major revision before its claims can be taken seriously. read the letter →

arxiv 2501.00779 v1 pith:VUMMTMMT submitted 2025-01-01 cs.SI cs.AI

Huyen Nguyen , Hieu Dam , Nguyen Do , Cong Tran , Cuong Pham This is my paper · ORCID
classification cs.SIcs.AI
keywords multiplexinfluencemaximizationspreadseedsetoptimizationreinforcementlearningvariationalautoencodermixtureofexpertsgraphneuralnetworks
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 claims that Multiplex Influence Maximization (MIM)—choosing a small seed set whose expected spread is largest in a multi-layer social network with heterogeneous diffusion models—can be solved by a fully learning-based pipeline that beats both combinatorial-optimization algorithms and prior deep/reinforcement-learning methods. The claim is empirical as well as architectural: on five real-world multiplex networks, under both Independent Cascade and Linear Threshold models, REM reports higher influence spread, better scalability, and faster inference than six baselines, with the largest margins on the biggest networks. The significance is practical: MIM is NP-hard, and traditional greedy methods time out or exhaust memory on million-node graphs, while earlier learning methods depend on expensive simulation or on high-quality training samples. REM attacks both bottlenecks by embedding seed sets in a continuous latent space and by replacing simulation with a learned mixture-of-experts spread estimator.

What carries the argument

The load-bearing object is the Propagation Mixture of Experts (PMoE) spread estimator $P(x, G; \xi)$: several GNN experts of different depths, with noisy top-$m$ routing, whose routed outputs are merged and normalized into a predicted infection count. PMoE does triple duty: it is the exploration objective in Eq. (11), the ranking score for the priority replay memory, and the objective for final gradient-ascent inference. Its companion is Seed2Vec, the VAE that maps discrete seed sets to a continuous latent space so that gradient methods can be applied at all. The theoretical glue is Theorem 3 (Influence Estimation Consistency), which is what authorizes replacing true spread with PMoE's predicted spread when optimizing $z$.

What would settle it

Run Algorithm 1's exploration on a multiplex network, record the latent samples $z$, and compare PMoE's predicted spread $P(D_\phi(z))$ with Monte Carlo spread $\delta(D_\phi(z))$ computed by Eq. (2); if the rank correlation is so low that the top-$k$ replay samples are not top-$k$ in true spread, or if replacing PMoE with true Monte Carlo spread as the selection and gradient score produces higher final influence, the central claim collapses.

Watch

Extended reading notes

Core claim

The central discovery, stated in the paper's Theorem 3, is that the discrete MIM search can be moved into a continuous latent space without changing the answer: once the VAE reconstruction error is minimized and the PMoE estimator is monotone, ordering seed sets by PMoE's predicted spread on latent codes is the same as ordering them by true spread in the original space, so the optimal seed set can be recovered by gradient ascent on the latent vector $z$. Around that identity, REM builds a loop: Seed2Vec encodes seed sets into a Gaussian latent space; an RL-style exploration objective pushes $z$ toward high-entropy, high-predicted-spread regions; the decoded candidates are ranked by PMoE, the top $k$ are replayed through a priority memory, and the model retrains on the augmented dataset. The paper's empirical assertion is that this loop, across Tables 1 to 3, yields the best reported spread and the shortest inference times among the compared methods.

Load-bearing premise

The load-bearing premise is that the PMoE spread predictor keeps its ranking faithful for latent seed sets it has never seen, so the seed sets it scores highest are also the ones with the largest true Monte Carlo spread.

Editorial extensions

If this is right

  • If REM is correct, inference-time seed selection no longer needs Monte Carlo simulation, which is what lets it handle graphs as large as the 1.9-million-node ParisAttack network where two greedy baselines run out of memory.
  • The same trained PMoE estimator serves every budget tested (1%, 5%, 10%, 20%) and both diffusion models, so one training pass covers a family of MIM instances rather than a single tuning.
  • The ablations imply the two added components are doing the work: dropping the RL exploration loop (REM-NonRL) and dropping the mixture-of-experts estimator (REM-NonMixture) both reduce final spread, so the reported gains are tied to those mechanisms.
  • If Theorem 3 holds in practice, then gradient ascent on the latent code $z$ is a valid substitute for discrete combinatorial search in MIM, not only on the tested networks.

Reading between the lines

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

  • A reader could test the framework's ceiling by replacing PMoE with Monte Carlo spread as the ranking score in the replay buffer: if true-spread replay yields better final seed sets, then REM's gains come from the surrogate's accuracy rather than from the exploration loop itself.
  • Because the paper's case study shows GAT experts beat GCN experts, the framework is modular: any future graph encoder could be dropped into the expert slots without changing the latent exploration, so REM's performance should track progress in graph representation learning.
  • The paper does not test whether PMoE's rankings survive when a deployment network follows a diffusion model different from the two used in training; in that setting the reported inference-time advantage could come at an unknown accuracy cost.
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 / 9 minor

Summary. The paper proposes REM, a learning-based framework for multiplex influence maximization (MIM). REM uses a VAE called Seed2Vec to embed seed sets into a continuous latent space, a Propagation Mixture of Experts (PMoE) composed of GNN experts to estimate influence spread, and an exploration objective (Eq. 11) that jointly maximizes latent entropy and PMoE-predicted spread. The generated samples are stored in a priority replay memory, and the top-k samples according to PMoE are combined with the original data to retrain the models. The paper claims that REM surpasses state-of-the-art methods in influence spread, scalability, and inference time, and it reports experiments on five multiplex networks under IC and LT diffusion models. The framework and the experimental setup are clearly described, but the central claims are not supported: the reported tables contradict the claimed consistent superiority, and the main theoretical guarantee (Theorem 3) rests on an invalid monotonicity lemma and a circular proof.

Significance. Multiplex influence maximization is an important problem, and the paper addresses it with a plausible architecture that combines latent-space optimization, a mixture-of-experts spread estimator, and experience replay. The experimental section is extensive in scope (five datasets, two diffusion models, seven baselines, and two ablations), and the authors state that code and datasets are available on GitHub. If the framework were correct and the claims properly qualified, it could be a useful contribution to the learning-based IM literature. However, the significance is substantially reduced because the paper's own tables contradict the abstract's claim of consistent superiority, and the theoretical justification for optimizing in latent space is not established. The self-referential use of PMoE as both the optimization objective and the replay-quality selector further raises correctness concerns that the paper does not resolve.

major comments (5)
  1. [Quantitative Analysis, Tables 1 and 2] The abstract and Section 'Quantitative Analysis' state that REM consistently outperforms other methods, but Tables 1 and 2 contain multiple cells where REM is strictly worse than baselines. In Table 1 (IC), ISF outperforms REM on Cora-ML at 1%, 5%, and 10% budgets (398.34 vs. 347.34; 778.62 vs. 765.48; 979.87 vs. 965.04) and on Celegans at 1% and 5% (1465.86 vs. 1445.16; 2298.01 vs. 2278.07), while MIM-Reasoner beats REM on Cora-ML at 1% (398.22 vs. 347.34). In Table 2 (LT), ISF, KSN, and MIM-Reasoner all beat REM on Cora-ML at every budget (e.g., 381 vs. 376 at 1% and 2145 vs. 2141 at 20%), and ISF beats REM on Celegans at 1% (1530 vs. 1514). These are not isolated or marginal differences; they appear systematically on smaller datasets and low-budget regimes. This is an internal inconsistency between the paper's stated claim and its reported evidence, and it undermines the central empirical contribution.
  2. [Appendix C1, Lemma 1] The proof of Lemma 1 is invalid. The proof asserts that 'softmax ... is non-decreasing', which is false: softmax is not component-wise monotone, because increasing one logit decreases the normalized probability of another. Even if each expert e_i were monotone in the seed set, the PMoE output in Eq. (9) is a convex combination with weights R_i(x) that themselves depend on x, so monotonicity of the combination does not follow. Additionally, the statement 'x(i) > x(j)' for binary vectors is undefined. Since Lemma 1 is invoked as the basis for the safety of optimizing the PMoE surrogate, this is a load-bearing gap in the theoretical argument.
  3. [Seed Node Set Inference and Appendix C3, Theorem 3] Theorem 3's proof is circular. It asserts that because reconstruction error is minimized, latent representations 'preserve the essential properties' of the original seed sets, and therefore if P on latent representations is ordered, the same ordering holds in the original space. This is exactly the claim of the theorem. Minimizing average MSE over a training set does not imply pairwise ordering preservation for arbitrary unseen latent pairs and their decoded seed sets; no formal argument is provided. Since Theorem 3 is the stated justification for the central method-level claim that 'the optimal seed set that maximizes influence can be found by optimizing z', the inference procedure lacks a valid correctness guarantee.
  4. [Latent Seed Set Exploration and Algorithm 1, Eq. (11)] The exploration objective in Eq. (11) uses the PMoE estimator P as a reward for generating new seed sets, and Algorithm 1 (lines 10-14) selects the top-k replay samples according to P and then retrains both Seed2Vec and PMoE on the combined dataset. This creates a self-referential loop in which P serves as both the optimization target and the quality selector for its own training data. The paper provides no argument, beyond the unproven Theorem 3, that systematic errors in P are corrected rather than amplified. This is a fundamental methodological concern because it affects the validity of the reported improvements and the reliability of the final seed set returned by Algorithm 2.
  5. [Scalability Analysis, Table 3] The inference-time claim is only partially supported by Table 3. At 10,000 nodes REM is slower than MIM-Reasoner (8.873s vs. 7.948s), and the text's statement that REM has 'on average, 10% faster inference time than the second-fastest MIM-Reasoner' is not backed by reported variance or any statistical test. Since inference time is one of the three claimed advantages in the abstract, the evidence for this part of the central claim is overstated.
minor comments (9)
  1. [Abstract] The phrase 'is a crucial as it can greatly diminish' is grammatically incomplete and should read 'is a crucial factor' or similar.
  2. [Conclusion and Appendix A] There are typos in 'Propagtaion' and 'ultilizes' that should be corrected.
  3. [Appendix D1, Table 4] The number of PMoE experts is listed as 8 in Table 4 but as 20 in the text of Appendix D1; this inconsistency needs to be resolved.
  4. [Appendix D1 vs. Algorithm 1] Appendix D1 states that Proximal Policy Optimization (PPO) is employed for policy training, but Algorithm 1 and Eq. (11) describe gradient descent on the latent variable z; the connection between PPO and the described algorithm is unexplained.
  5. [References] The reference to Stark et al. (2006) appears twice in identical form in the reference list and should be deduplicated.
  6. [Table 3] The header 'GCCOMB' is a typo for GCOMB.
  7. [Eq. (4)] The notation Fθ = Eψ ◦ Dϕ and the subsequent expression ˆx = Dϕ(Eψ(x)) are inconsistent with functional composition order; the intended order should be clarified.
  8. [Appendix C2, Lemma 2] The statement 'arg max_z H(Dϕ(z)) ∝ arg max_x H(x)' is not a well-defined proportionality between set-valued argmaxes; a formal definition of the relationship is needed.
  9. [Seed Node Set Inference] The sentence 'According to Theorem 3, the optimal seed set that maximizes influence can be found by optimizing z' overstates the theorem, which only claims an ordering preservation between latent and original spaces, not that the latent optimum corresponds to the global optimum of influence spread.

Circularity Check

2 steps flagged · score 7.0 of 10

REM's exploration loop labels and selects replay data with the same PMoE estimator it retrains, and Theorem 3's proof assumes the latent-to-original ordering it must establish.

  1. fitted input called prediction [Algorithm 1 (lines 8-14) and Section 'Latent Seed Set Exploration' (Eq. 11)]
    "Store (ˆx, P(Dϕ(z), G; ξ)) in PRM. ... Sample top k seed sets S (<k) t from PRM based on P. ... Retrain Fθ and P on combined dataset."

    The exploration loop stores, for each synthetic seed set, P's own prediction as the stored value; then it selects the top-k samples by that same P; then it retrains P on the combined dataset. For every synthetic sample, the PMoE loss (ˆy − y)2 is therefore computed with y = P_old(x̂), so retraining only asks P_new to reproduce P_old on inputs P_old already ranks highly. No ground-truth influence δ enters the loop for these samples, so the claimed 'reinforcement' and 'high-quality training samples' reduce to the model fitting its own prior. Any improvement over REM-NonRL attributed to this loop is, by construction, self-agreement rather than learning from the true influence signal.

  2. self definitional [Section 'Seed Node Set Inference' and Appendix C3 (Proof of Theorem 3)]
    "Since the Seed2Vec model minimizes reconstruction error, the latent representations z(i) and z(j) preserve the essential properties of their corresponding original seed sets x(i) and x(j). Therefore, if P(pϕ(z(i)), G; ξ) > P(pϕ(z(j)), G; ξ) in the latent space, the same ordering must hold in the original space, leading to P(x(i), G; ξ) > P(x(j), G; ξ)."

    Theorem 3 is the stated justification for Algorithm 2, which finds the seed set by gradient ascent on P(Dφ(z)) and then decodes. The proof's only substantive move is to assert that latent representations 'preserve the essential properties' and then conclude 'the same ordering must hold'—that is exactly the theorem to be proved. Lemma 1 gives monotonicity of P in the original seed-set space, not monotonicity of P with respect to latent perturbations, so it does not supply the missing step. Thus the claim that 'the optimal seed set that maximizes influence can be found by optimizing z' is, by construction, only a claim about optimizing the surrogate P; the reduction to true influence δ is assumed, not derived.

full rationale

The empirical headline—REM beats baselines on Monte-Carlo influence spread—is not circular: Tables 1-3 report externally simulated spread values, and the comparisons to MIM-Reasoner and other baselines are ordinary external benchmarks rather than load-bearing self-citations. The circularity lies in the internal derivation of how REM discovers and improves seed sets. First, the exploration loop of Eq. 11 and Algorithm 1 uses the PMoE estimator P as objective, reward, and label source: replay tuples store P's own value, top-k selection is by P, and P is retrained on the combined set, so the synthetic-data loss is the model fitting its own prior with no true influence signal. Second, the inference guarantee that optimizing z finds the optimal seed set rests on Theorem 3, whose proof simply asserts that reconstruction error makes latent-space ordering match original-space ordering; that assertion is the theorem itself. Algorithm 2 therefore returns the maximizer of P by construction, with the bridge to genuine influence spread δ assumed rather than demonstrated. Because these are central methodological claims, the score is elevated; because the final benchmark numbers are externally evaluated, the score is not a 9 or 10.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim rests on roughly six hand-picked or unreported parameters and four substantive assumptions. Two of those assumptions, Lemma 1 (PMoE monotonicity) and Theorem 3 (latent-to-original ordering preservation), are asserted with invalid or circular proofs, so they act as postulates rather than derived results. The other two, latent-space continuity of Seed2Vec and existence of a diverse initial dataset X0, are unquantified domain claims. The empirical evaluation itself is external (Monte Carlo simulation), which limits the circularity burden, but the method's internal improvement loop is self-referential.

free parameters (6)
  • KL weight = 0.55
    Hand-chosen weight balancing reconstruction and latent regularization in the Seed2Vec ELBO (Table 4); it shapes the latent space that exploration searches.
  • Entropy coefficient c = 0.1
    Hand-chosen coefficient scaling the entropy term in the exploration objective LExplore (Eq. 11 and Table 4).
  • Number of PMoE experts C = 8 in Table 4, 20 in Appendix D1
    Architecture choice that determines routing capacity; the paper contradicts itself on the value.
  • PMoE discretization threshold zeta = unspecified
    Used to convert predicted infection probabilities into discrete counts (Section 'Propagation Mixture of Expert'); its value is never reported.
  • Top-k count for replay sampling = unspecified
    Algorithm 1 samples 'top k seed sets' from the priority replay memory; k is never specified.
  • Latent dimension s = unspecified
    Dimension of the Seed2Vec latent variable z; never reported despite controlling the search space.
assumptions (6)
  • ad hoc to paper Monotonicity of the PMoE influence estimator (Lemma 1)
    Invoked as a 'theoretical guarantee' and used in the proof of Theorem 3. The supplied proof is invalid because softmax is not component-wise monotone and routing weights depend on x, so the lemma functions as an unproven assumption. Location: Appendix C1.
  • ad hoc to paper Reconstruction error minimization preserves the ordering of influence values between latent and original seed spaces (Theorem 3)
    The proof of Theorem 3 asserts this without derivation; it is the load-bearing step connecting optimization over z to optimization over x. Location: Appendix C3.
  • domain assumption A converged Seed2Vec guarantees latent-space continuity and completeness
    Stated in 'Latent Seed Set Exploration' as the reason exploration in z is meaningful; no quantitative argument is provided.
  • domain assumption Existence of a diverse initial seed-set dataset X0 with spread labels
    Seed2Vec and PMoE are initialized and periodically retrained on X0, and the paper cites data-quality dependence as a weakness it fixes, but never specifies how X0 is generated. Location: Definition 2 and Algorithm 1.
  • standard math Submodularity and monotonicity of influence spread for greedy guarantees
    Background result attributed to Kempe et al. 2003 and Kuhnle et al. 2018; not the paper's contribution.
  • standard math VAE and ELBO framework of Kingma and Welling 2013
    Standard generative modeling machinery used by Seed2Vec.

how reviews work

0 comments
Cite this review

Pith. "Pith review of REM: A Scalable Reinforced Multi-Expert Framework for Multiplex Influence Maximization." pith.science (2026). https://pith.science/paper/VUMMTMMT

@misc{pith2026250100779,
  author       = {Pith},
  title        = {Pith review of: REM: A Scalable Reinforced Multi-Expert Framework for Multiplex Influence Maximization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VUMMTMMT}},
  note         = {Machine review of arXiv:2501.00779}
}
read the original abstract

In social online platforms, identifying influential seed users to maximize influence spread is a crucial as it can greatly diminish the cost and efforts required for information dissemination. While effective, traditional methods for Multiplex Influence Maximization (MIM) have reached their performance limits, prompting the emergence of learning-based approaches. These novel methods aim for better generalization and scalability for more sizable graphs but face significant challenges, such as (1) inability to handle unknown diffusion patterns and (2) reliance on high-quality training samples. To address these issues, we propose the Reinforced Expert Maximization framework (REM). REM leverages a Propagation Mixture of Experts technique to encode dynamic propagation of large multiplex networks effectively in order to generate enhanced influence propagation. Noticeably, REM treats a generative model as a policy to autonomously generate different seed sets and learn how to improve them from a Reinforcement Learning perspective. Extensive experiments on several real-world datasets demonstrate that REM surpasses state-of-the-art methods in terms of influence spread, scalability, and inference time in influence maximization tasks.

Figures

Figures reproduced from arXiv: 2501.00779 by the authors.

Figure 1
Figure 1. An example illustrating the unique ”overlapping [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The diagram depicts REM’s process for addressing the MIM problem. Initially, REM utilizes Seed2Vec to embed [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Difference in influence spread (y-axis) of REM output on different dataset and budget when increasing exploration [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

74 extracted references · 68 canonical work pages

  1. [1]

    J. Zhu, J. Zhu, S. Ghosh, W. Wu, and J. Yuan. Social influence maximization in hypergraph in social networks. IEEE Transactions on Network Science and Engineering, 6(4):801--811, 2019

    work page 2019

  2. [2]

    Antelmi, G

    A. Antelmi, G. Cordasco, C. Spagnuolo, and P. Szufel. Social influence maximization in hypergraphs. Entropy, 23(7):796, 2021

    work page 2021

  3. [3]

    Influence Maximization in Hypergraphs

    M. Xie, X.-X. Zhan, C. Liu, and Z.-K. Zhang. Influence maximization in hypergraphs. arXiv preprint arXiv:2206.01394, 2022

    work page Pith review arXiv 2022

  4. [4]

    Wu and D

    J. Wu and D. Li. Modeling and maximizing information diffusion over hypergraphs based on deep reinforcement learning. Physica A: Statistical Mechanics and its Applications, 629:129193, 2023

    work page 2023

  5. [5]

    Banerjee, M

    S. Banerjee, M. Jenamani, and D. K. Pratihar. ComBIM: A community-based solution approach for the Budgeted Influence Maximization Problem. Expert Systems with Applications, 125:1--13, 2019

    work page 2019

  6. [6]

    S. S. Singh, A. Kumar, K. Singh, and B. Biswas. C2IM: Community based context-aware influence maximization in social networks. Physica A: Statistical Mechanics and its Applications, 514:796--818, 2019

    work page 2019

  7. [7]

    J. Zhu, S. Ghosh, W. Wu, and C. Gao. Profit maximization under group influence model in social networks. In International Conference on Computational Data and Social Networks, pages 108--119. Springer, 2019

    work page 2019

  8. [8]

    Banerjee and B

    S. Banerjee and B. Pal. Budgeted influence and earned benefit maximization with tags in social networks. Soc. Netw. Anal. Min., 12(1):21, 2022

    work page 2022

Show all 74 references
  1. [9]

    X. Chen, L. Deng, Y. Zhao, X. Zhou, and K. Zheng. Community-based influence maximization in location-based social network. WWWJ, 24(6):1903--1928, 2021

    1903

  2. [10]

    Contisciani, E

    M. Contisciani, E. A. Power, and C. De Bacco. Community detection with node attributes in multilayer networks. Scientific Reports, 2020

    2020

  3. [11]

    Kempe, J

    D. Kempe, J. Kleinberg, and \'E . Tardos. Maximizing the spread of influence through a social network. In Proc. of the KDD, 2003

    2003

  4. [12]

    Leskovec, A

    J. Leskovec, A. Krause, C. Guestrin, C. Faloutsos, J. VanBriesen, and N. Glance. Cost-effective outbreak detection in networks. In Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 420--429, 2007

    2007

  5. [13]

    Goyal, W

    A. Goyal, W. Lu, and L. Lakshmanan. CELF++: Optimizing the greedy algorithm for influence maximization in social networks. In Proceedings of the 20th International Conference Companion on World Wide Web, WWW 2011, pages 47--48, 2011

    2011

  6. [14]

    Y. Tang, X. Xiao, and Y. Shi. Influence maximization: Near-optimal time complexity meets practical efficiency. In Proc. of the SIGMOD, pages 75--86, 2014

    2014

  7. [15]

    Kuhnle, M

    A. Kuhnle, M. A. Alim, X. Li, H. Zhang, and M. T. Thai. Multiplex influence maximization in online social networks with heterogeneous diffusion models. IEEE Transactions on Computational Social Systems, 5(2):418--429, 2018

    2018

  8. [16]

    S. Bai, F. Zhang, and P. H. S. Torr. Hypergraph convolution and hypergraph attention. Pattern Recognit., 110:107637, 2021

    2021

  9. [17]

    Q. Li, Z. Han, and X.-M. Wu. Deeper insights into graph convolutional networks for semi-supervised learning. In AAAI Conference on Artificial Intelligence, 2018

    2018

  10. [18]

    N. M. Shazeer, A. Mirhoseini, K. Maziarz, A. Davis, Q. V. Le, G. E. Hinton, and J. Dean. Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. In International Conference on Learning Representations, 2017

    2017

  11. [19]

    B. W. Dolhansky and J. A. Bilmes. Deep submodular functions: Definitions and learning. In Advances in Neural Information Processing Systems, volume 29, 2016

    2016

  12. [20]

    D. P. Kingma and M. Welling. Auto-encoding variational bayes. In International Conference on Learning Representations, 2013

    2013

  13. [21]

    T. Chen, S. Yan, J. Guo, and W. Wu. ToupleGDD: A fine-designed solution of influence maximization by deep reinforcement learning. arXiv preprint arXiv:2210.07500, 2022

    2022 arXiv

  14. [22]

    Manchanda, A

    S. Manchanda, A. Mittal, A. Dhawan, S. Medya, S. Ranu, and A. Singh. GCOMB: Learning budget-constrained combinatorial algorithms over billion-sized graphs. In Advances in Neural Information Processing Systems, pages 20000--20011, 2020

    2020

  15. [23]

    Y. Li, G. Wang, X. Ji, Y. Xiang, and D. Fox. DeepIM: Deep iterative matching for 6D pose estimation. International Journal of Computer Vision, 2018

    2018

  16. [24]

    N. H. K. Do, T. Chowdhury, C. Ling, L. Zhao, and M. T. Thai. MIM-Reasoner: Learning with theoretical guarantees for multiplex influence maximization. In International Conference on Artificial Intelligence and Statistics, pages 2296--2304, 2024

    2024

  17. [25]

    J. S. Lim, S. Y. Ri, B. D. Egan, and F. A. Biocca. The cross-platform synergies of digital video advertising: Implications for cross-media campaigns in television, Internet and mobile TV. Computers in Human Behavior, 48:463--472, 2015

    2015

  18. [26]

    Q. Zhan, J. Zhang, S. Wang, P. S. Yu, and J. Xie. Influence maximization across partially aligned heterogenous social networks. In Pacific-Asia conference on knowledge discovery and data mining, pages 58--69. Springer, 2015

    2015

  19. [27]

    Zhang, D

    H. Zhang, D. T. Nguyen, H. Zhang, and M. T. Thai. Least cost influence maximization across multiple social networks. IEEE/ACM Transactions on Networking, 24(2):929--939, 2016

    2016

  20. [28]

    S. S. Singh, K. Singh, A. Kumar, and B. Biswas. MIM2: Multiple influence maximization across multiple social networks. Physica A: Statistical Mechanics and its Applications, 526:120902, 2019

    2019

  21. [29]

    Katukuri, M

    M. Katukuri, M. Jagarapu, and others. CIM: clique-based heuristic for finding influential nodes in multilayer networks. Applied Intelligence, 52(5):5173--5184, 2022

    2022

  22. [30]

    Domingos and M

    P. Domingos and M. Richardson. Mining the network value of customers. In Proceedings of the Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 57--66. ACM, 2001

    2001

  23. [31]

    W. O. Kermack and A. G. McKendrick. A contribution to the mathematical theory of epidemics. Proceedings of the royal society of london. Series A, Containing papers of a mathematical and physical character, 115(772):700--721, 1927

    1927

  24. [32]

    Stark, B.-J

    C. Stark, B.-J. Breitkreutz, T. Reguly, L. Boucher, A. Breitkreutz, and M. Tyers. BioGRID: A general repository for interaction datasets. Nucleic acids research, 34:D535--9, 2006

    2006

  25. [33]

    Omodei, M

    E. Omodei, M. D. De Domenico, and A. Arenas. Characterizing interactions in online social networks during exceptional events. Frontiers in Physics, 3, 2015

    2015

  26. [34]

    De Domenico and E

    M. De Domenico and E. G. Altmann. Unraveling the origin of social bursts in collective attention. Scientific reports, 10(1):4629, 2020

    2020

  27. [35]

    C. Zhou, P. Zhang, W. Zang, and L. Guo. On the upper bounds of spread for greedy algorithms in social network influence maximization. IEEE Transactions on Knowledge and Data Engineering, 27(10):2770--2783, 2015

    2015

  28. [36]

    Kimura and K

    M. Kimura and K. Saito. Tractable models for information diffusion in social networks. In Knowledge Discovery in Databases: PKDD 2006, pages 259--271. Springer Berlin Heidelberg, 2006

    2006

  29. [37]

    W. Chen, Y. Yuan, and L. Zhang. Scalable influence maximization in social networks under the linear threshold model. In 2010 IEEE International Conference on Data Mining, pages 88--97. IEEE, 2010 a

    2010

  30. [38]

    W. Chen, C. Wang, and Y. Wang. Scalable influence maximization for prevalent viral marketing in large-scale social networks. In Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 1029--1038. ACM, 2010 b

    2010

  31. [39]

    Y. Li, J. Fan, Y. Wang, and K.-L. Tan. Influence maximization on social graphs: A survey. TKDE, 30(10):1852--1872, 2018

    2018

  32. [40]

    Y. Tang, Y. Shi, and X. Xiao. Influence maximization in near-linear time: A martingale approach. In Proc. of the SIGMOD, 2015

    2015

  33. [41]

    J. Tang, X. Tang, X. Xiao, and J. Yuan. Online processing algorithms for influence maximization. In Proc. of the SIGMOD, pages 991--1005, 2018

    2018

  34. [42]

    Q. Guo, S. Wang, Z. Wei, and M. Chen. Influence maximization revisited: Efficient reverse reachable set generation with bound tightened. In Proc. of the SIGMOD, pages 2167--2181, 2020

    2020

  35. [43]

    H. T. Nguyen, M. T. Thai, and T. N. Dinh. Stop-and-stare: Optimal sampling algorithms for viral marketing in billion-scale networks. In Proc. of the SIGMOD, 2016

    2016

  36. [44]

    Jiang, G

    Q. Jiang, G. Song, C. Gao, Y. Wang, W. Si, and K. Xie. Simulated annealing based influence maximization in social networks. In Proc. of the AAAi, 2011

    2011

  37. [45]

    X. Li, J. D. Smith, T. N. Dinh, and M. T. Thai. Tiptop:(almost) exact solutions for influence maximization in billion-scale networks. IEEE/ACM Transactions on Networking, 27(2):649--661, 2019

    2019

  38. [46]

    Banerjee, M

    S. Banerjee, M. Jenamani, and D. K. Pratihar. A survey on influence maximization in a social network. KAIS, 62(9):3417--3455, 2020

    2020

  39. [47]

    Lin, S.-D

    S.-C. Lin, S.-D. Lin, and M.-S. Chen. A learning-based framework to handle multi-round multi-party influence maximization on social networks. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 695--704, 2015

    2015

  40. [48]

    Ali, C.-Y

    K. Ali, C.-Y. Wang, and Y.-S. Chen. Boosting reinforcement learning in competitive influence maximization with transfer learning. In 2018 IEEE/WIC/ACM International Conference on Web Intelligence (WI), pages 395--400. IEEE, 2018

    2018

  41. [49]

    H. Li, M. Xu, S. S. Bhowmick, C. Sun, Z. Jiang, and J. Cui. Disco: Influence maximization meets network embedding and deep learning. arXiv preprint arXiv:1906.07378, 2019

    1906 arXiv

  42. [50]

    C. Ling, J. Jiang, J. Wang, and L. Zhao. Source localization of graph diffusion via variational autoencoders for graph inverse problems. In Proc. of the KDD, 2022

    2022

  43. [51]

    Manchanda, A

    S. Manchanda, A. Mittal, A. Dhawan, S. Medya, S. Ranu, and A. Singh. Gcomb: Learning budget-constrained combinatorial algorithms over billion-sized graphs. Advances in Neural Information Processing Systems, 33:20000--20011, 2020

    2020

  44. [52]

    Cai and Y

    C. Cai and Y. Wang. A note on over-smoothing for graph neural networks. arXiv preprint arXiv: 2006.13318, 2020

    2006 arXiv

  45. [53]

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

    2022

  46. [54]

    C. Ling, T. Chowdhury, J. Jiang, J. Wang, X. Zhang, H. Chen, and L. Zhao. DeepGAR: Deep graph learning for analogical reasoning. In 2022 IEEE International Conference on Data Mining (ICDM), pages 1065--1070. IEEE, 2022

    2022

  47. [55]

    C. Ling, J. Jiang, J. Wang, M. T. Thai, R. Xue, J. Song, M. Qiu, and L. Zhao. Deep graph representation learning and optimization for influence maximization. In International Conference on Machine Learning, pages 21350--21361. PMLR, 2023

    2023

  48. [56]

    Chowdhury, C

    T. Chowdhury, C. Ling, J. Jiang, J. Wang, M. T. Thai, and L. Zhao. Deep graph representation learning influence maximization with accelerated inference. Available at SSRN 4663083, 2024

    2024

  49. [57]

    Zhang, W

    C. Zhang, W. Li, D. Wei, Y. Liu, and Z. Li. Network dynamic GCN influence maximization algorithm with leader fake labeling mechanism. IEEE Transactions on Computational Social Systems, 2022

    2022

  50. [58]

    Kumar, A

    S. Kumar, A. Mallik, A. Khetarpal, and B. S. Panda. Influence maximization in social networks using graph embedding and graph neural network. Information Sciences, 607:1617--1636, 2022

    2022

  51. [59]

    S. Tian, S. Mo, L. Wang, and Z. Peng. Deep reinforcement learning-based approach to tackle topic-aware influence maximization. Data Science and Engineering, 5(1):1--11, 2020

    2020

  52. [60]

    Y. Lin, W. Chen, and J. C. Lui. Boosting information spread: An algorithmic approach. In 2017 IEEE 33rd International Conference on Data Engineering (ICDE), pages 883--894. IEEE, 2017

    2017

  53. [61]

    Z. Yuan, M. Shao, and Z. Chen. Graph bayesian optimization for multiplex influence maximization. In Thirty-Eighth AAAI Conference on Artificial Intelligence, AAAI 2024, pages 22475--22483. AAAI Press, 2024

    2024

  54. [62]

    Liaghat, A

    Z. Liaghat, A. Rasekh, and A. Mahdavi. Application of data mining methods for link prediction in social networks. Social Network Analysis and Mining, 3(1):97--111, 2013

    2013

  55. [63]

    D. F. Nettleton. Data mining of social networks represented as graphs. Computer Science Review, 7:1--34, 2013

    2013

  56. [64]

    F. Bonchi. Influence propagation in social networks: A data mining perspective. In 2011 IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology, 1:2--2. IEEE, 2011

    2011

  57. [65]

    Vikatos, P

    P. Vikatos, P. Gryllos, and C. Makris. Marketing campaign targeting using bridge extraction in multiplex social network. Artificial Intelligence Review, 53:2335--2360, 2020

    2020

  58. [66]

    Zhang, X

    H. Zhang, X. Chen, Y. Peng, G. Kou, and R. Wang. The interaction of multiple information on multiplex social networks. Information Sciences, 605:366--380, 2022

    2022

  59. [67]

    Jalili, Y

    M. Jalili, Y. Orouskhani, M. A. Mehrabadi, N. Alipourfard, and M. Perc. Link prediction in multiplex online social networks. Royal Society Open Science, 4(2):160863, 2017

    2017

  60. [68]

    A. K. McCallum, K. Nigam, J. Rennie, and K. Seymore. Automating the construction of internet portals with machine learning. Information Retrieval, 3(2):127--163, 2000

    2000

  61. [69]

    Schaul, J

    T. Schaul, J. Quan, I. Antonoglou, and D. Silver. Prioritized experience replay. In International Conference on Learning Representations, 2015

    2015

  62. [70]

    Horgan, J

    D. Horgan, J. Quan, D. Budden, G. Barth-Maron, M. Hessel, H. Van Hasselt, and D. Silver. Distributed prioritized experience replay. arXiv preprint arXiv:1803.00933, 2018

    2018 arXiv

  63. [71]

    D. P. Kingma. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013

    2013 arXiv

  64. [72]

    H. M. Ngo, N. H. K. Do, M. N. Vu, T. Kahveci, and M. T. Thai. CHARME: A chain-based reinforcement learning approach for the minor embedding problem. arXiv preprint arXiv:2406.07124, 2024

    2024

  65. [73]

    , " * 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...

  66. [74]

    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 10, 2026 · model on record in the stance chip above.