Pith. sign in

REVIEW 2 major objections 6 minor 36 references

HyperDet: Source Detection in Hypergraphs via Interactive Relationship Construction and Feature-rich Attention Fusion

T0 review · 2 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper reports the first source-detection method designed for hypergraphs, with F1 gains of 8–15 percent over propagation-aware baselines and 23–30 percent over centrality/user-state methods across eight datasets.

desk verdict The first hypergraph source-detection paper with a usable architecture, but the evaluation may leak the answer through exact infection timestamps in the input features. read the letter →

arxiv 2505.12894 v2 pith:RDOBETRY submitted 2025-05-19 cs.SI cs.AI

Le Cheng , Peican Zhu , Yangming Guo , Keke Tang , Chao Gao , Zhen Wang This is my paper · ORCID
classification cs.SIcs.AI
keywords hypergraphrumorsourcedetectionlocalizationhigher-orderinteractionsgraphneuralnetworksattentionmechanisminformationpropagationindependentcascademodel
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

Rumor source detection has mostly assumed information spreads through pairwise contacts, but in classrooms, households, and online groups people influence one another in larger clusters. This paper claims to be the first to formalize source detection on hypergraphs—networks whose edges can contain many users at once—and proposes HyperDet to solve it. HyperDet builds a node-relationship representation from both the static hypergraph structure and the live split between spreaders and ignorants, then learns node representations through an autoencoder and a multi-head attention mechanism that emphasizes nodes with strong transmission roles. On eight datasets with simulated hypergraph cascades, the paper reports F1-score gains of 8–15 percent over propagation-aware baselines and 23–30 percent over centrality- and user-state-based baselines. If those numbers hold, group structure is not something to flatten into pairwise edges but a resource for pinpointing where a rumor began.

What carries the argument

The load-bearing construction is the augmented incidence matrix $$H' = H \parallel H_{\mathrm{ig}} \parallel H_{\mathrm{sp}},$$ which folds three kinds of relationship into one hypergraph: the original static hyperedges, one hyperedge containing every current spreader, and one containing every ignorant node. The raw node feature $X_i$ concatenates a state label (+1 for spreader, −1 otherwise), the propagation timestamp (or −1 for ignorants), and the $k$ smallest nontrivial Laplacian eigenvectors of the infected subgraph as positional encoding. These inputs feed the Feature-rich Attention Fusion module: a hypergraph autoencoder whose layers alternate node-to-hyperedge and hyperedge-to-node attention convolution, followed by multi-head self-attention that gives different nodes different weight, with the loss balanced by $\rho = |s|/(n - |s|)$. The whole pipeline is what lets the model learn node representations under higher-order relations instead of pairwise flattening.

What would settle it

Simulate hypergraph cascades with group-propagation probability set to zero ($p_\Delta = 0$) while keeping every other setting identical, and compare HyperDet against GIN-SD; if the reported 8–15 percent F1 advantage disappears or reverses, the paper's attribution of the gains to higher-order interaction modeling is refuted.

Watch

Extended reading notes

Core claim

The central claim, stated the way the authors would state it, is that a source detector which models higher-order interactions directly, rather than flattening them into pairwise edges, can locate rumor origins in hypergraphs more accurately than pairwise methods. HyperDet takes a single snapshot of an evolving hypergraph: its topology, each node's spreader/ignorant state, and propagation timestamps, and outputs a predicted source set. The Interactive Relationship Construction module concatenates the static incidence matrix $H$ with two dynamic hyperedges $H_{\mathrm{ig}}$ (all ignorants) and $H_{\mathrm{sp}}$ (all spreaders), and composes raw features from state, timestamp, and Laplacian positional encoding of the infected subgraph. The Feature-rich Attention Fusion module autoencodes these features through hypergraph attention convolution, reweights nodes by multi-head self-attention, and uses a class-balancing coefficient to counter the rarity of sources. Across the authors' experiments on eight datasets, HyperDet achieves the best accuracy, F1-score, and AUC on every dataset.

Load-bearing premise

The load-bearing premise is that the synthetic hypergraph cascade used for evaluation—with individual spread probability drawn uniformly from $[0, 0.5]$ and group pressure set to $p_\Delta = 0.3\,(|e \cap G^+|/|e|)$—captures the way rumors actually move through real group interactions.

Editorial extensions

If this is right

  • Pairwise source detectors cannot simply be copied onto hypergraphs: the paper shows their performance degrades when group interactions are present, whereas a detector that models the group structure directly keeps high F1.
  • Early detection works: as the rumor scale grows from 10% to 30%, HyperDet remains ahead of all baselines on every dataset, so the model does not need a saturated network to identify the source.
  • It tolerates missing data: when up to 25% of nodes lack features, HyperDet's F1 declines less than the baselines, because structural positional encoding still carries information.
  • The architecture is not tied to one diffusion model: stable results are reported under SI, SIS, SIR, and independent-cascade dynamics, so the two-module design generalizes across propagation assumptions.

Reading between the lines

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

  • Beyond the paper: the decisive test is real-world deployment. Because all reported cascades are simulated with one specific group-pressure formula, we should expect the 8–15 percent edge to shrink or shift when applied to real message-forwarding data with ground-truth origins.
  • If higher-order interaction modeling is truly the cause of the gains, then sweeping the group-propagation probability $p_\Delta$ downward from 0.3 toward 0 should make HyperDet's advantage over pairwise baselines shrink monotonically; this is a cheap, direct check of the paper's main explanatory story.
  • The idea of encoding the current spreader/ignorant partition as two extra hyperedges is portable. It could be applied to other dynamic hypergraph tasks—for instance, predicting which node will be affected next, or detecting coordinated groups—where the current infection state is a conditioning variable rather than the target.
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

2 major / 6 minor

Summary. The paper introduces HyperDet, a supervised method for source detection in hypergraphs. It builds an Interactive Relationship Construction (IRC) module that combines the static hypergraph topology with dynamic state-based hyperedges, and a Feature-rich Attention Fusion (FAF) module that applies a hypergraph autoencoder and multi-head attention to learned node features, together with a class-balanced loss. The method is evaluated on eight datasets under a synthetic hypergraph independent cascade model, reporting consistent gains over centrality-, user-state-, and propagation-based baselines, along with ablations, early-detection analyses, and robustness checks under alternative diffusion models.

Significance. If the evaluation is valid, HyperDet would be a meaningful contribution: it defines a hypergraph-specific source-detection setting, and the reported F1 improvements of roughly 8-15% over propagation-based baselines are substantial. The paper also includes a broad set of datasets, module ablations, and sensitivity-style experiments. However, the central claim is not currently established because the raw input features appear to contain exact infection timestamps that make the synthetic source-detection task nearly trivial, and because the evaluation lacks repeated trials, error bars, and statistical testing. The paper does not release code, which further limits reproducibility. The strength of the contribution therefore depends on resolving the timestamp-leakage concern.

major comments (2)
  1. [Section 4.1, Eq. (5); Section 5.1] The raw feature X_i^2 defined in Eq. (5) is the exact infection timestamp t_i for every spreader node and -1 for ignorant nodes. In the synthetic IC model of Section 5.1, 5% of nodes are selected as sources and become spreaders at t=0, while all other infections occur at later steps. Consequently, the set {v_i : X_i^2 = 0} is exactly the ground-truth source set, and a trivial threshold or ranker on this single feature solves the task perfectly without any hypergraph modeling, giving F1 and AUC near 1.0. The paper does not report a timestamp-only baseline, does not state whether the baselines were given the same timestamp feature, and does not explain why HyperDet's reported F1 is only 0.63-0.80 and AUC 0.86-0.92 under these conditions. If timestamps were available to HyperDet but not to baselines, the comparison is unfair; if they were hidden, noised, or unavailable, that must be stated explicitly. Please remove the exact timestamp from the input, use noised or censored timestamps, add a timestamp-only or minimum-timestamp baseline, and report all methods under identical feature availability. This is necessary before the claimed 8-15% improvement can be attributed to higher-order relationship modeling rather than to feature leakage.
  2. [Section 5.1, Table 1] All results in Table 1 appear to be single runs with no error bars, no standard deviations, and no significance tests. The evaluation also uses one fixed synthetic propagation model (p ~ U(0,0.5), p_delta = 0.3(|e ∩ G+|/|e|), 5% sources, 30% snapshot threshold), and the datasets are not actual social networks but standard UCI/benchmark datasets with the synthetic process superimposed. Under this protocol, the reported performance gaps between methods may reflect run-to-run variability rather than model quality, and the conclusions cannot be extrapolated to other propagation dynamics. Please report means and standard deviations over multiple random seeds, include a statistical significance test (e.g., paired t-test or Wilcoxon signed-rank test), and provide sensitivity analyses over the propagation model parameters. The alternative-model results in Table 4 should also be compared with baselines, not only with HyperDet.
minor comments (6)
  1. [Section 4.1, Eq. (8)] The positional feature X_i^3 is a k-dimensional vector for spreader nodes but is written as the scalar -1 for ignorant nodes; if the concatenated raw feature is to have a fixed dimension, the -1 case needs to be defined as a vector of -1s of the same length, or otherwise standardized.
  2. [Section 4.2, Eq. (12)] There is a dimension mismatch in the attention coefficient formula: W is defined in R^{l'_w x l_w}, but it is applied to the concatenated vector [X_i || theta_j], which has length 2l_w; the dimensions of W or of the concatenation need to be corrected for the model to be implementable as written.
  3. [Section 5.3, Table 3] The sentence 'w/o A achieves the best results' is ambiguous because the table shows that the full HyperDet model outperforms w/o A; please clarify that w/o A is the best among the three attention variants (w/o A, w/ AL, w/ AS) and that the full attention fusion still contributes positively.
  4. [Section 5.1] The experimental settings refer to 'small networks (G1-G4)', but the datasets are named Zoo, House, NTU2012, and Mushroom; please align the notation with the dataset names used in Table 1.
  5. [Section 1, Contributions] The claim 'We are the first to formalize the problem of source detection in hypergraphs' should be softened or supported by a more thorough literature search, since the paper does not discuss any prior hypergraph-based source localization work and the current survey in Section 2 stops at hypergraph representation learning.
  6. [Section 5.2, Table 4] Table 4 reports HyperDet under SI, SIS, SIR, and IC models, but it does not include any baseline method under the same alternative models; adding at least the strongest baseline (e.g., GIN-SD) would make the robustness claim more convincing.

Circularity Check

1 steps flagged · score 6.0 of 10

Exact infection-timestamp feature (Eq. 5) makes the source set recoverable by thresholding in the synthetic IC model, so the reported gains do not validate the hypergraph modules.

  1. self definitional [Section 4.1 Node Feature Construction, Eq. (5); Section 3 Propagation Process on Hypergraphs; Section 5.1 Experimental Settings]
    "X2 i = ti, v i ∈ G + ; −1, otherwise (Eq. 5). ... at t= 0, the set of source nodes s in the network transitions from the ignorant (susceptible) state to the spreader (infected) state. ... Initially, 5% of the nodes are selected as sources ... A network snapshot is captured when 30% of the nodes become spreaders."

    In the synthetic IC model used for evaluation, all sources are exactly the spreader nodes with infection time t=0; every non-source spreader has t>0. Eq. (5) feeds the exact infection time t_i as raw feature X_i^2 for every spreader node. Therefore, within the spreader set, the ground-truth source set is precisely the set of nodes attaining the minimum value of X_i^2, recoverable by a trivial rank or threshold on this single feature, with no need for hypergraph structure, IRC, autoencoder, or attention. The paper does not state that baselines received the same timestamp feature, and no timestamp-only baseline is reported.

full rationale

HyperDet is otherwise an empirical supervised pipeline with no derivation of outcomes from inputs; the IRC/FAF modules are standard learnable components, the loss uses true labels only during training, and the comparison includes external baselines plus the authors' GIN-SD as a legitimate benchmark. The load-bearing circular step is confined to the evaluation protocol: Eq. (5) encodes the exact infection time, while the generative model defines sources as the t=0 nodes. Hence the target is a deterministic function of the raw feature, making the experimental superiority claim not self-contained evidence for the proposed hypergraph mechanism. No other reductions by construction, ansatz-smuggling citations, or load-bearing self-citations were found.

Assumptions & free parameters 6 free parameters · 3 assumptions · 0 invented entities

The central claim rests on several hand-chosen simulation parameters and domain assumptions about how rumors spread on hypergraphs. None of these are fitted to real propagation data, and the paper provides no evidence that the chosen generation model matches real-world dynamics.

free parameters (6)
  • group propagation coefficient pΔ = 0.3
    pΔ = 0.3(|e∩G+|/|e|) sets the strength of group influence in the synthetic propagation model. This value is chosen by hand and directly shapes the simulated data on which all methods are evaluated.
  • per-node propagation probability p = U(0, 0.5)
    Each node's pairwise propagation probability is sampled from a uniform distribution, defining the synthetic rumor dynamics.
  • source ratio = 5%
    5% of nodes are selected as sources in every simulation. This arbitrary choice sets the class imbalance and task difficulty.
  • snapshot threshold = 30%
    A snapshot is captured when 30% of nodes become spreaders, determining the observation point for source identification.
  • latent feature dimension = 64
    The autoencoder latent dimension is fixed at 64 for all networks, a hyperparameter that affects capacity.
  • attention heads and hidden neurons = varies by network size
    Small networks use 3 heads and 500 neurons, medium networks use 2 heads and 500 neurons, and the large network uses 1 head and 400 neurons. These architectural choices are not justified.
assumptions (3)
  • domain assumption The hypergraph independent cascade model with group pressure pΔ = 0.3(|e∩G+|/|e|) adequately represents rumor propagation in social networks.
    All evaluation data is generated by this model, so any superiority of HyperDet is only demonstrated under this synthetic dynamics.
  • domain assumption Applying pairwise-graph baseline methods to the clique expansion of hypergraphs is a fair comparison.
    Baselines are designed for pairwise graphs; adapting them via clique expansion is assumed to be an appropriate and representative baseline.
  • domain assumption The selected datasets (Zoo, House, NTU2012, Mushroom, ModelNet40, 20News, PubMed, Walmart) are representative of social networks with higher-order interactions.
    Several are not social networks (Zoo is an animal taxonomy, ModelNet40 is 3D objects), yet the paper draws social-network conclusions from them without justification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HyperDet: Source Detection in Hypergraphs via Interactive Relationship Construction and Feature-rich Attention Fusion." pith.science (2026). https://pith.science/paper/RDOBETRY

@misc{pith2026250512894,
  author       = {Pith},
  title        = {Pith review of: HyperDet: Source Detection in Hypergraphs via Interactive Relationship Construction and Feature-rich Attention Fusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RDOBETRY}},
  note         = {Machine review of arXiv:2505.12894}
}
read the original abstract

Hypergraphs offer superior modeling capabilities for social networks, particularly in capturing group phenomena that extend beyond pairwise interactions in rumor propagation. Existing approaches in rumor source detection predominantly focus on dyadic interactions, which inadequately address the complexity of more intricate relational structures. In this study, we present a novel approach for Source Detection in Hypergraphs (HyperDet) via Interactive Relationship Construction and Feature-rich Attention Fusion. Specifically, our methodology employs an Interactive Relationship Construction module to accurately model both the static topology and dynamic interactions among users, followed by the Feature-rich Attention Fusion module, which autonomously learns node features and discriminates between nodes using a self-attention mechanism, thereby effectively learning node representations under the framework of accurately modeled higher-order relationships. Extensive experimental validation confirms the efficacy of our HyperDet approach, showcasing its superiority relative to current state-of-the-art methods.

Figures

Figures reproduced from arXiv: 2505.12894 by the authors.

Figure 1
Figure 1. Social scenario modeling. (a) Various social scenarios. (b) [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Framework of HyperDet. (a) The snapshot acquisition process. (b) The Interactive Relationship Construction (IRC) module, which [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Visualization of source detection results on House. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: The performance in early rumor sources detection. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Impact of data incompleteness on source detection. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 33 canonical work pages

  1. [1]

    EPA: Exoneration and prominence based age for infection source identification

    [Aliet al., 2019 ] Syed Shafat Ali, Tarique Anwar, Ajay Ras- togi, and Syed Afzal Murtaza Rizvi. EPA: Exoneration and prominence based age for infection source identification. InProceedings of the 28th ACM International Conference on Information and Knowledge Management, pages 891– 900,

    work page 2019

  2. [7]

    On visual similarity based 3d model retrieval

    [Chenet al., 2003 ] Ding-Yun Chen, Xiao-Pei Tian, Yu-Te Shen, and Ming Ouhyoung. On visual similarity based 3d model retrieval. InComputer Graphics Forum, volume 22, pages 223–232. Wiley Online Library,

    work page 2003

  3. [8]

    Efficient source detection in incomplete networks via sensor deployment and source approaching.IEEE Transactions on Information Forensics and Security,

    [Chenget al., 2025 ] Le Cheng, Peican Zhu, Keke Tang, Chao Gao, and Zhen Wang. Efficient source detection in incomplete networks via sensor deployment and source approaching.IEEE Transactions on Information Forensics and Security,

    work page 2025

  4. [11]

    Social contagion models on hypergraphs.Physical Review Research, 2(2):023032,

    [de Arrudaet al., 2020 ] Guilherme Ferraz de Arruda, Gio- vanni Petri, and Yamir Moreno. Social contagion models on hypergraphs.Physical Review Research, 2(2):023032,

    work page 2020

  5. [12]

    Multiple ru- mor source detection with graph convolutional networks

    [Donget al., 2019 ] Ming Dong, Bolong Zheng, Nguyen Quoc Viet Hung, Han Su, and Guohui Li. Multiple ru- mor source detection with graph convolutional networks. InProceedings of the 28th ACM International Conference on Information and Knowledge Management, pages 569– 578,

    work page 2019

  6. [13]

    Wavefront- based multiple rumor sources identification by multi-task learning.IEEE Transactions on Emerging Topics in Com- putational Intelligence, 6(5):1068–1078,

    [Donget al., 2022 ] Ming Dong, Bolong Zheng, Guohui Li, Chenliang Li, Kai Zheng, and Xiaofang Zhou. Wavefront- based multiple rumor sources identification by multi-task learning.IEEE Transactions on Emerging Topics in Com- putational Intelligence, 6(5):1068–1078,

    work page 2022

  7. [14]

    Unveiling implicit deceptive patterns in multi-modal fake news via neuro-symbolic reasoning

    [Donget al., 2024 ] Yiqi Dong, Dongxiao He, Xiaobao Wang, Youzhu Jin, Meng Ge, Carl Yang, and Di Jin. Unveiling implicit deceptive patterns in multi-modal fake news via neuro-symbolic reasoning. InProceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 8354–8362,

    work page 2024

  8. [15]

    Hypergraph neural networks

    [Fenget al., 2019 ] Yifan Feng, Haoxuan You, Zizhao Zhang, Rongrong Ji, and Yue Gao. Hypergraph neural networks. InProceedings of the AAAI Conference on Ar- tificial Intelligence, volume 33, pages 3558–3565,

    work page 2019

Show all 36 references
  1. [16]

    HGNN+: General hypergraph neural networks

    [Gaoet al., 2022 ] Yue Gao, Yifan Feng, Shuyi Ji, and Ron- grong Ji. HGNN+: General hypergraph neural networks. IEEE Transactions on Pattern Analysis and Machine In- telligence, 45(3):3181–3199,

    2022

  2. [17]

    Unignn: a unified framework for graph and hypergraph neural net- works

    [Huang and Yang, 2021] Jing Huang and Jie Yang. Unignn: a unified framework for graph and hypergraph neural net- works. InIJCAI, pages 2563–2569,

    2021

  3. [18]

    Enhancing multi-scale diffusion prediction via sequential hypergraphs and adver- sarial learning

    [Jiaoet al., 2024 ] Pengfei Jiao, Hongqian Chen, Qing Bao, Wang Zhang, and Huaming Wu. Enhancing multi-scale diffusion prediction via sequential hypergraphs and adver- sarial learning. InProceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 8571–8581,

    2024

  4. [19]

    Local-global de- fense against unsupervised adversarial attacks on graphs

    [Jinet al., 2023 ] Di Jin, Bingdao Feng, Siqi Guo, Xiaobao Wang, Jianguo Wei, and Zhen Wang. Local-global de- fense against unsupervised adversarial attacks on graphs. InProceedings of the AAAI Conference on Artificial Intel- ligence, volume 37, pages 8105–8113,

    2023

  5. [22]

    Source localization of graph diffusion via variational autoencoders for graph inverse problems

    [Linget al., 2022 ] Chen Ling, Junji Jiang, Junxiang Wang, and Zhao Liang. Source localization of graph diffusion via variational autoencoders for graph inverse problems. InProceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 1010– 1020,

    2022

  6. [23]

    Locating the source of diffusion in large-scale networks.Physical Review Letters, 109(6):068702,

    [Pintoet al., 2012 ] Pedro C Pinto, Patrick Thiran, and Mar- tin Vetterli. Locating the source of diffusion in large-scale networks.Physical Review Letters, 109(6):068702,

    2012

  7. [25]

    Rumors in a network: Who’s the culprit?IEEE Transac- tions on Information Theory, 57(8):5163–5181,

    [Shah and Zaman, 2011] Devavrat Shah and Tauhid Zaman. Rumors in a network: Who’s the culprit?IEEE Transac- tions on Information Theory, 57(8):5163–5181,

    2011

  8. [27]

    Graph attention networks

    [Veliˇckovi´cet al., 2017 ] Petar Veliˇckovi´c, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Lio, and Yoshua Bengio. Graph attention networks. InInterna- tional Conference on Learning Representations,

    2017

  9. [28]

    Multiple source detection without knowing the underlying propagation model

    [Wanget al., 2017 ] Zheng Wang, Chaokun Wang, Jisheng Pei, and Xiaojun Ye. Multiple source detection without knowing the underlying propagation model. InProceed- ings of the AAAI Conference on Artificial Intelligence, vol- ume 31,

    2017

  10. [29]

    An invertible graph diffusion neural network for source localization

    [Wanget al., 2022 ] Junxiang Wang, Junji Jiang, and Liang Zhao. An invertible graph diffusion neural network for source localization. InProceedings of the ACM Web Con- ference 2022, pages 1058–1069,

    2022

  11. [30]

    Elevating knowledge-enhanced entity and relation- ship understanding for sarcasm detection.IEEE Transac- tions on Knowledge and Data Engineering,

    [Wanget al., 2025 ] Xiaobao Wang, Yujing Wang, Dongxiao He, Zhe Yu, Yawen Li, Longbiao Wang, Jianwu Dang, and Di Jin. Elevating knowledge-enhanced entity and relation- ship understanding for sarcasm detection.IEEE Transac- tions on Knowledge and Data Engineering,

    2025

  12. [31]

    3d shapenets: A deep representation for volumetric shapes

    [Wuet al., 2015 ] Zhirong Wu, Shuran Song, Aditya Khosla, Fisher Yu, Linguang Zhang, Xiaoou Tang, and Jianxiong Xiao. 3d shapenets: A deep representation for volumetric shapes. InProceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition, pages 1912–1920,

    2015

  13. [33]

    Locating the propaga- tion source in complex networks with a direction-induced search based gaussian estimator.Knowledge-Based Sys- tems, 195:105674,

    [Yanget al., 2020 ] Fan Yang, Shuhong Yang, Yong Peng, Yabing Yao, Zhiwen Wang, Houjun Li, Jingxian Liu, Ruisheng Zhang, and Chungui Li. Locating the propaga- tion source in complex networks with a direction-induced search based gaussian estimator.Knowledge-Based Sys- tems, 19...

    2020

  14. [34]

    Learning with hypergraphs: Clus- tering, classification, and embedding.Advances in Neural Information Processing Systems, 19,

    [Zhouet al., 2006 ] Dengyong Zhou, Jiayuan Huang, and Bernhard Sch ¨olkopf. Learning with hypergraphs: Clus- tering, classification, and embedding.Advances in Neural Information Processing Systems, 19,

    2006

  15. [36]

    Locating multi-sources in social networks with a low infection rate.IEEE Transactions on Network Science and Engineering, 9(3):1853–1865, 2022

    [Zhuet al., 2022 ] Peican Zhu, Le Cheng, Chao Gao, Zhen Wang, and Xuelong Li. Locating multi-sources in social networks with a low infection rate.IEEE Transactions on Network Science and Engineering, 9(3):1853–1865, 2022

    2022

  16. [2006]

    Catch’em all: Locating multiple diffusion sources in net- works with partial observations

    [Zhuet al., 2017 ] Kai Zhu, Zhen Chen, and Lei Ying. Catch’em all: Locating multiple diffusion sources in net- works with partial observations. InProceedings of the AAAI Conference on Artificial Intelligence, volume 31,

    2017

  17. [2007]

    Hypergraph convolution and hypergraph attention

    [Baiet al., 2021 ] Song Bai, Feihu Zhang, and Philip HS Torr. Hypergraph convolution and hypergraph attention. Pattern Recognition, 110:107637,

    2021

  18. [2011]

    Finding patient zero: Learn- ing contagion source with graph neural networks.arXiv preprint arXiv:2006.11913,

    [Shahet al., 2020 ] Chintan Shah, Nima Dehmamy, Nicola Perra, Matteo Chinazzi, Albert-L ´aszl´o Barab´asi, Alessan- dro Vespignani, and Rose Yu. Finding patient zero: Learn- ing contagion source with graph neural networks.arXiv preprint arXiv:2006.11913,

    2020 arXiv

  19. [2012]

    Spotting culprits in epidemics: How many and which ones? In2012 IEEE 12th International Conference on Data Mining, pages 11–20

    [Prakashet al., 2012 ] B Aditya Prakash, Jilles Vreeken, and Christos Faloutsos. Spotting culprits in epidemics: How many and which ones? In2012 IEEE 12th International Conference on Data Mining, pages 11–20. IEEE,

    2012

  20. [2015]

    Hypergcn: A new method for training graph convolutional networks on hypergraphs

    [Yadatiet al., 2019 ] Naganand Yadati, Madhav Nimishakavi, Prateek Yadav, Vikram Nitin, Anand Louis, and Partha Talukdar. Hypergcn: A new method for training graph convolutional networks on hypergraphs. Advances in Neural Information Processing Systems, 32,

    2019

  21. [2017]

    Propagation source identification of infec- tious diseases with graph convolutional networks.Journal of Biomedical Informatics, 116:103720,

    [Liet al., 2021 ] Liang Li, Jianye Zhou, Yuewen Jiang, and Biqing Huang. Propagation source identification of infec- tious diseases with graph convolutional networks.Journal of Biomedical Informatics, 116:103720,

    2021

  22. [2019]

    Clustering in graphs and hypergraphs with cate- gorical edge labels

    [Amburget al., 2020 ] Ilya Amburg, Nate Veldt, and Austin Benson. Clustering in graphs and hypergraphs with cate- gorical edge labels. InProceedings of The Web Conference 2020, pages 706–717,

    2020

  23. [2020]

    Uci machine learning repository,

    [Asuncionet al., 2007 ] Arthur Asuncion, David Newman, et al. Uci machine learning repository,

    2007

  24. [2021]

    Graph contrastive learning for source localization in social networks.Information Sci- ences, page 121090,

    [Baoet al., 2024 ] Qing Bao, Ying Jiang, Wang Zhang, Pengfei Jiao, and Jing Su. Graph contrastive learning for source localization in social networks.Information Sci- ences, page 121090,

    2024

  25. [2022]

    Generative hypergraph clustering: From blockmodels to modularity.Science Advances, 7(28):eabh1303,

    [Chodrowet al., 2021 ] Philip S Chodrow, Nate Veldt, and Austin R Benson. Generative hypergraph clustering: From blockmodels to modularity.Science Advances, 7(28):eabh1303,

    2021

  26. [2023]

    Semi-supervised classification with graph convolutional networks

    [Kipf and Welling, 2017] Thomas N Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In5th International Conference on Learning Representations, ICLR,

    2017

  27. [2024]

    Networks beyond pairwise interactions: Structure and dynamics

    [Battistonet al., 2020 ] Federico Battiston, Giulia Cencetti, Iacopo Iacopini, Vito Latora, Maxime Lucas, Alice Pata- nia, Jean-Gabriel Young, and Giovanni Petri. Networks beyond pairwise interactions: Structure and dynamics. Physics Reports, 874:1–92,

    2020

  28. [2025]

    You are allset: A multiset function framework for hypergraph neural networks

    [Chienet al., 2022 ] Eli Chien, Chao Pan, Jianhao Peng, and Olgica Milenkovic. You are allset: A multiset function framework for hypergraph neural networks. InICLR,

    2022

Pith tools

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