Multi-view Fake News Detection Model Based on Dynamic Hypergraph

Rongping Ye; Xiaobing Pei

REVIEW 3 major objections 6 minor 27 references

Multi-view Fake News Detection Model Based on Dynamic Hypergraph

T0 review · 3 major / 6 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 A multi-view fake news detector that dynamically rewires a news hypergraph during convolution, and aligns propagation-tree and hypergraph embeddings by authenticity, reports the best accuracy and F1 among nine baselines on PolitiFact and…

desk verdict A sensible three-view fake news detector whose headline numbers rest on an unstated train/test mask in the contrastive loss; fix that and the paper is a solid incremental contribution. read the letter →

arxiv 2412.19227 v1 pith:PYGD6D4R submitted 2024-12-26 cs.LG

Rongping Ye , Xiaobing Pei This is my paper · ORCID

classification cs.LG

keywords fakenewsdetectiondynamichypergraphstructurelearningmulti-viewcontrastivepropagationtreesocialmedia

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes a fake news detection model that combines three views of a news item: the text, its propagation tree on social media, and a hypergraph linking news pieces that share users, publication times, or entities. The key move is to treat the hypergraph structure as learnable: after each convolution layer the model rebuilds hyperedges from node and hyperedge embedding similarities, so the high-order relationships are optimized for the task rather than fixed in advance. It also adds supervised contrastive learning between the propagation and hypergraph views, pulling same-authenticity news together and pushing opposite-authenticity news apart. On PolitiFact and Gossipcop the model reports the best accuracy and F1 among nine baselines, with accuracy 92.81% and 98.84% respectively.

What carries the argument

The load-bearing mechanism is a dynamic hypergraph convolution block. At each layer, after nodes pass messages to hyperedges, a weighted cosine similarity matrix between news embeddings and hyperedge embeddings selects, per hyperedge, the top-$p_{thd}$ fraction of most similar nodes, producing a reconstructed incidence matrix $H_{re}$; a graph-level attention layer then fuses the original hypergraph with the reconstructed one and a text-derived hypergraph. A supervised InfoNCE contrastive loss between propagation-tree embeddings $X_{pro}$ and hypergraph embeddings $X_{hg}$ pulls same-label news together and pushes different-label news apart, making the learned embeddings authenticity-relevant.

What would settle it

Train DHy-MFND with the contrastive loss in Eq. (16) restricted to the training split, exactly as the loss must be if labels are not to leak; if the resulting PolitiFact or Gossipcop accuracy falls materially below the reported 92.81% or 98.84%, the reported superiority over baselines is not explained by the described method alone.

Watch

Extended reading notes

Core claim

The paper claims that fake news detection improves when the relational structure used to compare news pieces is itself learned from the data, rather than fixed by handcrafted rules. DHy-MFND starts with a predefined hypergraph whose hyperedges group news by shared users, close publication times, or similar entities, then, during hypergraph convolution, recomputes these hyperedges from the current node and hyperedge embeddings using weighted cosine similarity, fuses the original and rebuilt structures with a graph-level attention layer, and repeats this rewiring at every layer. At the same time, a supervised InfoNCE loss aligns the propagation-tree embedding and hypergraph embedding of news with the same authenticity label and separates pairs with different labels. The fused multi-view embedding is classified by a softmax. The reported consequence is that this design beats nine baselines on both datasets, with the largest gains on PolitiFact, where hypergraph-based relational reasoning helps most.

Load-bearing premise

The method's headline results assume that validation and test labels are never used to build the training signal; the paper writes the contrastive loss over all news pieces using ground-truth labels and never states that the loss excludes the validation and test splits.

Editorial extensions

If this is right

The reported results imply that hand-built hyperedges are a floor, not a ceiling: letting the structure adapt during training improves both accuracy and F1 over the fixed-hypergraph baseline on both datasets.
Removing the propagation-tree view degrades the model more than removing the hypergraph view on Gossipcop, while the reverse holds on PolitiFact, so the three views carry complementary signal rather than redundancy.
Because the model needs only news text, propagation trees, and user interaction data, it can be deployed where large heterogeneous graphs with external knowledge bases are unavailable.
The ablation without contrastive learning drops about four accuracy points on PolitiFact, so the supervised alignment between the two views is a substantial contributor, not a marginal trick.

Reading between the lines

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

Editorial inference: the dynamic structure-learning block is not news-specific; it should transfer to other hypergraph node-classification problems where initial hyperedges are noisy, such as rumor stance detection or document classification.
Editorial inference: the threshold $p_{thd}$ acts as a hard top-k selection, so replacing it with a differentiable soft assignment would let the structure learning be trained end-to-end with straight-through gradients, a direct variant the paper does not explore.
Editorial inference: the paper's runtime table shows Gossipcop takes 311.73 seconds per epoch versus 7.51 seconds for PolitiFact; scaling the dynamic rewiring to larger hypergraphs would require subsampling or incremental hyperedge updates, which the paper does not address.

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.

Desk Editor's Note

read the letter

The paper is a clean engineering combination: text (BERT), propagation tree (GraphSAGE), and hypergraph (attention-based HGNN) views, with dynamic hypergraph structure learning borrowed from HSL and HGFND's hypergraph construction. The idea of optimizing the hypergraph while learning embeddings, plus InfoNCE between propagation and hypergraph views, is coherent and worth testing. The paper is clearly written and the architecture is reproducible in principle. It does not break new conceptual ground, but it is a reasonable step forward for a crowded subfield.

The main problem is the contrastive loss in Eq. (16). It is written as a sum over all N news pieces, with positive and negative sets defined by ground-truth labels. Algorithm 1 takes all labels as input and never restricts the loss to the training split. With a 6:2:2 split and a hypergraph built over all news nodes, this can leak validation/test labels into training. The paper never states that K(i) and T(i) are masked to training indices. That is not a minor typo; it directly affects whether Table 3's numbers mean anything. The stress-test note is right to flag it as the load-bearing uncertainty.

Other soft spots are smaller. The ablation study (Table 5) reports no variance, even though the experiment section says all models were run 20 times and means with standard deviations are reported. That makes it hard to judge whether differences like w/o HG vs w/o Pro are real. The PolitiFact dataset has only 314 news pieces, and the claimed gain over FinerFact (92.81±2.50 vs 91.48±1.89) is within one standard deviation; the Gossipcop gain over HGFND is more convincing. Hyperparameter analysis covers only p_thd, not lambda or tau, and the chosen values for those are not given.

The paper cites HSL and HGFND properly, and the combination is new enough for an incremental paper. The central idea is not flawed; the execution has a reproducibility gap that can be fixed by rewriting the loss with an explicit training mask and reporting ablations with variance.

I would not cite the numbers as they stand, but I would bring this to a reading group as a case study in how transductive settings can blur label leakage in contrastive learning. A serious editor should send it to peer review, asking the authors to clarify the masking and provide code/data. If the leakage is resolved, this is a publishable incremental contribution.

Referee Report

3 major / 6 minor

Summary. The paper proposes DHy-MFND, a multi-view fake news detection model that combines BERT-based text embeddings, a GraphSAGE propagation-tree encoder, and an attention-based hypergraph neural network, together with a dynamic hypergraph structure learning (DHSL) module that refines the incidence matrix during training. A supervised InfoNCE contrastive loss aligns propagation-tree and hypergraph embeddings for news with the same authenticity label. Experiments on PolitiFact and Gossipcop report accuracy and F1 scores over nine baselines, with the proposed model achieving the highest mean scores in Table 3. The paper also includes an ablation study, a runtime table, and a sensitivity analysis of the threshold ratio p_thd.

Significance. If the reported results are valid, the paper makes a modest but useful contribution by showing that combining text, propagation-tree, and hypergraph views with dynamic hypergraph structure learning and supervised contrastive learning can improve fake news detection on two standard benchmarks. The paper is generally reproducible in structure: it provides algorithm pseudocode, reports means and standard deviations for the main comparisons, and includes a hyperparameter sensitivity analysis. The central empirical claim, however, hinges on the correct isolation of training labels in the contrastive loss and on the statistical robustness of the reported gains, both of which need clarification.

major comments (3)

[Eq. (16), Algorithm 1] The concern raised in the stress-test about label leakage is well-founded. The contrastive loss in Eq. (16) is written as a sum over all N news pieces, with K(i) = {k : y_k = y_i} and T(i) = {t : y_t ≠ y_i} using ground-truth labels. The dataset is split 6:2:2, but the hypergraph and all computations are described on the full set of N nodes. Algorithm 1 takes the full label vector Y as input and computes the contrastive loss without any visible restriction to the training split. If validation or test labels participate in the positive/negative pair construction during training, the reported Acc/F1 numbers in Table 3 are invalid. Please state explicitly whether a training mask is applied; if not, rerun the experiments with the contrastive loss computed only on training-set nodes and report whether the headline results still hold.
[Eq. (15)] The cross-entropy loss in Eq. (15) is given as L_ce = -1/N Σ Σ ŷ_{i,c} log(y_{i,c}), with y_{i,c} defined as the ground-truth label and ŷ_{i,c} as the predicted probability. As written, the ground-truth label appears inside the logarithm, which is not a valid classification loss and would involve log(0) for the zero class. Please correct the equation to the standard form (ground truth outside the log, predicted probability inside) and confirm that the implementation matches the corrected form, since Eq. (17) combines this loss with the contrastive loss.
[Table 3] The claim that DHy-MFND 'consistently outperforms' the baselines is not fully supported by the reported variance. On PolitiFact, the mean accuracy gain over FinerFact is 92.81±2.50 vs 91.48±1.89, a difference within one standard deviation, and no significance tests are reported. On Gossipcop, the gain over HGFND is 98.84±0.30 vs 97.46±0.30, which is more substantial, but the overall claim needs statistical backing. Please report paired significance tests (e.g., t-test or Wilcoxon signed-rank over the 20 runs) and provide standard deviations for the ablation results in Table 5, which are currently reported without any variance.

minor comments (6)

[Eq. (14), Algorithm 1] In Eq. (14) the reconstructed incidence matrix is referred to as both H_re and H_rec, and Algorithm 1 calls 'Generate H_re using Eq. (13)' twice; please disambiguate the update of H^{(l+1)} in the DHSL recursion.
[Figure 4] The sensitivity plot in Figure 4 does not label the y-axis; please state which metric (Acc or F1) is shown and make the axis labels explicit.
[Baselines] The paper says 9 baselines, but Table 3 lists three UPFD variants (GCN, GAT, SAGE) in addition to the other methods; please clarify whether UPFD is counted as one or three separate baselines in the comparison.
[Table 2 and Table 3] The dataset name is spelled 'Politifact' in Table 2 and 'PolitiFact' in Table 3 and elsewhere; please standardize the spelling.
[Experimental Setup] The paper does not describe how the validation split is used (e.g., early stopping or hyperparameter selection); please specify this to rule out any selection of the best epoch or test configuration on the test set.
[Hyperparameter Analysis] In the Hyperparameter Analysis section, the sentence 'For 0 ≤ p_thd ≤ 1, we conduct repeated experiments...' appears twice; please remove the duplicate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical training-and-evaluation study whose components are standard and whose claims are benchmarked against external baselines.

full rationale

This is an empirical training-and-evaluation paper. The reported contributions—BERT-based text encoding, GraphSAGE propagation-tree encoding, attention-based HGNN encoding, dynamic hypergraph structure learning, and InfoNCE-based contrastive learning—are standard building blocks, and the paper's central claims are evaluated against nine external baselines on two benchmark datasets. No result in the paper is derived from a fitted constant or from an equation that is equivalent to its own input by construction. The DHSL module optimizes the hypergraph structure using node and hyperedge embeddings, but this optimization is part of the end-to-end supervised training objective; it is not a circular 'prediction' of the classification outcome. The contrastive loss in Eq. (16) uses ground-truth labels to define positive and negative pairs, which is standard supervised contrastive learning and not a circularity, though it raises a separate potential label-leakage concern because the loss is written over all N news pieces while the dataset is split 6:2:2. That concern is a correctness and validity question about whether the implementation masks validation/test labels, not a circularity of the kind this pass targets, and the paper text alone does not prove leakage. The paper also contains no self-citation: the authors, Ye and Pei, do not appear in the reference list, and no load-bearing premise is justified by a self-citation chain. The ablation studies and hyperparameter analysis are ordinary empirical analyses. Overall, the derivation chain is self-contained in the sense relevant to circularity: every component is either independently defined or externally benchmarked, and no claim reduces to its own inputs by construction. Therefore the circularity score is 0.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The model's central claim rests on the quality of the predefined hypergraph, the supervised contrastive objective, and the unstated assumption that labels are only used for training nodes. The free parameters (p_thd, lambda, tau, hypergraph thresholds) are not fully reported, limiting independent verification.

free parameters (4)

p_thd (similarity threshold ratio) = Not reported; swept from 0 to 1 in steps of 0.1 in Figure 4
Controls the number of nodes in newly generated hyperedges in the DHSL module. The chosen value is not stated in the text.
lambda (contrastive loss weight) = Not reported
Balances cross-entropy and InfoNCE losses in Eq. (17). Its value is presumably tuned but never given.
tau (InfoNCE temperature) = Not reported
Temperature parameter in the contrastive loss (Eq. 16). Not specified.
Hypergraph construction thresholds = Not reported
The predefined hypergraph is inherited from HGFND, but the thresholds for same-user, same-time, and similar-entity hyperedges are not provided.

assumptions (4)

domain assumption Hyperedges based on same user, same publication time, and similar entities capture high-order relationships useful for fake news detection.
Invoked in the Hypergraph-level View and in the definition of the predefined hypergraph structure following HGFND (Jeong et al. 2022). No independent evidence is given for the informativeness of these specific hyperedge types.
domain assumption Supervised contrastive learning with label-defined positive and negative pairs improves authenticity-relevant news embeddings.
Invoked in the Contrastive Learning section (Eq. 16). This is a working hypothesis, not proven in the paper.
domain assumption BERT-encoded news text and user profile features (from UPFD) are sufficient initial node features for the propagation and hypergraph views.
Used throughout the model as X_text and X_usr; the paper does not test alternative feature encoders.
domain assumption The contrastive loss is computed only on training-set nodes, despite being written over all N news pieces with ground-truth labels.
The paper splits data 6:2:2, and the hypergraph contains all news nodes. If the loss uses validation/test labels, results would be invalid. The paper does not state this restriction.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Multi-view Fake News Detection Model Based on Dynamic Hypergraph." pith.science (2026). https://pith.science/paper/PYGD6D4R

@misc{pith2026241219227,
  author       = {Pith},
  title        = {Pith review of: Multi-view Fake News Detection Model Based on Dynamic Hypergraph},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PYGD6D4R}},
  note         = {Machine review of arXiv:2412.19227}
}

read the original abstract

With the rapid development of online social networks and the inadequacies in content moderation mechanisms, the detection of fake news has emerged as a pressing concern for the public. Various methods have been proposed for fake news detection, including text-based approaches as well as a series of graph-based approaches. However, the deceptive nature of fake news renders text-based approaches less effective. Propagation tree-based methods focus on the propagation process of individual news, capturing pairwise relationships but lacking the capability to capture high-order complex relationships. Large heterogeneous graph-based approaches necessitate the incorporation of substantial additional information beyond news text and user data, while hypergraph-based approaches rely on predefined hypergraph structures. To tackle these issues, we propose a novel dynamic hypergraph-based multi-view fake news detection model (DHy-MFND) that learns news embeddings across three distinct views: text-level, propagation tree-level, and hypergraph-level. By employing hypergraph structures to model complex high-order relationships among multiple news pieces and introducing dynamic hypergraph structure learning, we optimize predefined hypergraph structures while learning news embeddings. Additionally, we introduce contrastive learning to capture authenticity-relevant embeddings across different views. Extensive experiments on two benchmark datasets demonstrate the effectiveness of our proposed DHy-MFND compared with a broad range of competing baselines.

Figures

Figures reproduced from arXiv: 2412.19227 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 6 canonical work pages

[3]

IEEE Transactions on Pattern Analysis and Machine Intelligence 44(5): 2548-2566

Hypergraph Learning: Methods and Practices. IEEE Transactions on Pattern Analysis and Machine Intelligence 44(5): 2548-2566. doi.org/10.1109/tpami.2020.3039374. Yang, Y.; Huang, C.; Xia, L.; Liang, Y.; Yu, Y.; and Li, C

arXiv 2020
[6]

In Proceedings of the Twenty- Eighth International Joint Conference on Artificial Intelligence , 2635-2641

Dynamic Hypergraph Neural Networks. In Proceedings of the Twenty- Eighth International Joint Conference on Artificial Intelligence , 2635-2641. doi.org/10.24963/ijcai.2019/366. Wei, L.; Hu, D.; Lai, Y.; Zhou, W.; and Hu, S

work page doi:10.24963/ijcai.2019/366 2019
[9]

Community Detection in Large Hypergraphs

Community detection in large hypergraphs. arXiv:2301.11226. Wang, C.; Wang, X.; Li, Z.; Chen, Z.; and Li, J

work page Pith review arXiv
[10]

In Proceedings of the AAAI Conference on Artificial Intelli- gence, 5746–5754

Towards Fine-Grained Reasoning for Fake News Detec- tion. In Proceedings of the AAAI Conference on Artificial Intelli- gence, 5746–5754. doi.org/10.1609/aaai.v36i5.20517. Lu, Y. Ju.; and Li, C.T

work page doi:10.1609/aaai.v36i5.20517
[11]

arXiv:2004.11648

GCAN: Graph-Aware Co-Attention Networks for Explainable Fake News Detection on Social Media. arXiv:2004.11648. Bian, T.; Xiao, X.; Xu, T.; Zhao, P.; Huang, W.; Rong, Y.; and Huang, J

arXiv 2004
[12]

In Proceedings of the AAAI Conference on Artificial Intelligence, 549–556

Rumor Detection on Social Media with Bi-Direc- tional Graph Convolutional Networks. In Proceedings of the AAAI Conference on Artificial Intelligence, 549–556. doi.org/10.1609/aaai.v34i01.5393. Dou, Y.; Shu, K.; Xia, C.; Yu, P.S.; and Sun, L

doi:10.1609/aaai.v34i01.5393
[13]

In Proceedings of the 44th International ACM SIGIR Conference on Research and Develop- ment in Information Retrieval, 2051-2055

User Pref- erence-Aware Fake News Detection. In Proceedings of the 44th International ACM SIGIR Conference on Research and Develop- ment in Information Retrieval, 2051-2055. doi.org/10.1145/3404835.3462990. Su, X.; Yang, J.; Wu, J.; and Zhang, Y

arXiv
[14]

In Proceedings of the sixteenth ACM interna- tional conference on web search and data mining , 51-59

Mining User-Aware Multi-Relations for Fake News Detection in Large Scale Online Social Networks. In Proceedings of the sixteenth ACM interna- tional conference on web search and data mining , 51-59. doi.org/10.1145/3539597.3570478. Zhang, J.; Chen, Y.; Xiao, X.; Lu, R.; and Xia, S

arXiv

Show all 27 references

[15]

arXiv:2106.06666

Learnable hypergraph laplacian for hypergraph learning. arXiv:2106.06666. Cai, D.; Song, M.; Sun, C.; Zhang, B.; Hong, S.; and Li, H

arXiv
[16]

In IJCAI ,1923-1929

Hypergraph Structure Learning for Hypergraph Neural Networks. In IJCAI ,1923-1929. He, Z.; Li, C.; Zhou, F.; and Yang, Y

1923
[17]

In Proceedings of the 44th International ACM SIGIR Conference on Research and De- velopment in Information Retrieval, 2020-2024

Rumor Detection on Social Media with Event Augmentations. In Proceedings of the 44th International ACM SIGIR Conference on Research and De- velopment in Information Retrieval, 2020-2024. doi.org/10.1145/3404835.3463001. Sun, T.; Qian, Z.; Dong, S.; Li, P.; and Zhu, Q

2020
[18]

In Proceedings of the ACM Web Conference , 2789 -2797

Rumor De- tection on Social Media with Graph Adversarial Contrastive Learn- ing. In Proceedings of the ACM Web Conference , 2789 -2797. doi.org/10.1145/3485447.3511999. Oord, A.; Li, Y.; and Vinyals, O
[19]

arXiv:1902.06673

Fake news detection on social media using geometric deep learning. arXiv:1902.06673. Han, Y.; Karunasekera, S.; and Leckie, C

1902 arXiv
[20]

In Proceedings of the International AAAI Conference on Web and Social Media, 626–637

Hierar- chical Propagation Networks for Fake News Detection: Investiga- tion and Exploitation. In Proceedings of the International AAAI Conference on Web and Social Media, 626–637. doi.org/10.1609/icwsm.v14i1.7329. Nguyen, V.H.; Sugiyama, K.; Nakov, P.; and Kan, M.Y

doi
[21]

In Proceedings of the 29th ACM interna- tional conference on information & knowledge management , 1165-1174

Fang: Leveraging social context for fake news detection using graph representation . In Proceedings of the 29th ACM interna- tional conference on information & knowledge management , 1165-1174. doi.org/10.1145/3340531.3412046. Cui, J.; Kim, K.; Na, S.H.; and Shin, S
[22]

In Pro- ceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2253-2262

Reinforcement subgraph reasoning for fake news detection. In Pro- ceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2253-2262. doi.org/10.1145/3534678.3539277. Shu, K.; Mahudeswaran, D.; Wang, S.; Lee, D.; and Liu, H
[23]

Big Data 8(3): 171–188

FakeNewsNet: A Data Repository with News Content, Social Con- text, and Spatiotemporal Information for Studying Fake News on Social Media. Big Data 8(3): 171–188. doi.org/10.1089/big.2020.0062. Jeong, U.; Ding, K.; Cheng, L.; Guo, R.; Shu, K.; and Liu, H. 2022 Nothing stands a...

2020
[24]

arXiv:1810.04805

Bert: Pre- training of deep bidirectional transformers for language under- standing. arXiv:1810.04805. Huang, J.; and Yang, J

arXiv
[25]

arXiv:2105.00956

Unignn: a unified framework for graph and hypergraph neural networks. arXiv:2105.00956. Chien, E. ; Pan, C.; Peng, J.; and Milenkovic, O

arXiv
[26]

arXiv:2106.13264

You are allset: A multiset function framework for hypergraph neural net- works. arXiv:2106.13264. Bai, S.; Zhang, F.; and Torr, P.H.S

arXiv
[27]

Pattern Recognition , 110, 107637

Hypergraph Convolution and Hypergraph Attention. Pattern Recognition , 110, 107637. doi.org/10.1016/j.patcog.2020.107637. Yun, S.; Jeong, M.; Kim, R.; Kang, J.; and Kim, H.J

2020
[2018]

arXiv:1807.03748

Representation Learning with Contrastive Predictive Coding. arXiv:1807.03748. Shu, K.; Mahudeswaran, D.; Wang, S.; and Liu, H

arXiv
[2019]

In Proceedings of the AAAI Conference on Artificial Intelligence, 3558–3565

Hyper - graph Neural Networks. In Proceedings of the AAAI Conference on Artificial Intelligence, 3558–3565. doi.org/10.1609/aaai.v33i01.33013558. Jiang, J.; Wei, Y.; Feng, Y.; Cao, J.; and Gao, Y

doi
[2020]

arXiv: 2007.03316

Graph neural net- works with continual learning for fake news detection from social media. arXiv: 2007.03316. Wu, H.; Li, N.; Zhang, J.; Chen, S.; Ng, M.K.; Long, J

2007 arXiv
[2021]

In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, 3347-3363

STANKER: Stack- ing Network Based on Level-Grained Attention-Masked BERT for Rumor Detection on Social Media. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, 3347-3363. doi.org/10.18653/v1/2021.emnlp-main.269. Gong, S.; Sinnott, R.O.;...

2021 doi
[2022]

In Proceedings of the 28th ACM SIGKDD Con- ference on Knowledge Discovery and Data Mining , 2263 -2274

Multi-Behavior Hypergraph-Enhanced Transformer for Sequential Recommendation. In Proceedings of the 28th ACM SIGKDD Con- ference on Knowledge Discovery and Data Mining , 2263 -2274. doi.org/10.1145/3534678.3539342. Feng, Y.; You, H.; Zhang, Z.; Ji, R.; and Gao, Y
[2023]

arXiv:2307.12639

Fake news de- tection through graph -based neural networks: A survey. arXiv:2307.12639. Gao, Y.; Zhang, Z.; Lin, H.; Zhao, X.; Du, S.; and Zou, C

arXiv

Pith tools

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

[1] [3]

IEEE Transactions on Pattern Analysis and Machine Intelligence 44(5): 2548-2566

Hypergraph Learning: Methods and Practices. IEEE Transactions on Pattern Analysis and Machine Intelligence 44(5): 2548-2566. doi.org/10.1109/tpami.2020.3039374. Yang, Y.; Huang, C.; Xia, L.; Liang, Y.; Yu, Y.; and Li, C

arXiv 2020

[2] [6]

In Proceedings of the Twenty- Eighth International Joint Conference on Artificial Intelligence , 2635-2641

Dynamic Hypergraph Neural Networks. In Proceedings of the Twenty- Eighth International Joint Conference on Artificial Intelligence , 2635-2641. doi.org/10.24963/ijcai.2019/366. Wei, L.; Hu, D.; Lai, Y.; Zhou, W.; and Hu, S

work page doi:10.24963/ijcai.2019/366 2019

[3] [9]

Community Detection in Large Hypergraphs

Community detection in large hypergraphs. arXiv:2301.11226. Wang, C.; Wang, X.; Li, Z.; Chen, Z.; and Li, J

work page Pith review arXiv

[4] [10]

In Proceedings of the AAAI Conference on Artificial Intelli- gence, 5746–5754

Towards Fine-Grained Reasoning for Fake News Detec- tion. In Proceedings of the AAAI Conference on Artificial Intelli- gence, 5746–5754. doi.org/10.1609/aaai.v36i5.20517. Lu, Y. Ju.; and Li, C.T

work page doi:10.1609/aaai.v36i5.20517

[5] [11]

arXiv:2004.11648

GCAN: Graph-Aware Co-Attention Networks for Explainable Fake News Detection on Social Media. arXiv:2004.11648. Bian, T.; Xiao, X.; Xu, T.; Zhao, P.; Huang, W.; Rong, Y.; and Huang, J

arXiv 2004

[6] [12]

In Proceedings of the AAAI Conference on Artificial Intelligence, 549–556

Rumor Detection on Social Media with Bi-Direc- tional Graph Convolutional Networks. In Proceedings of the AAAI Conference on Artificial Intelligence, 549–556. doi.org/10.1609/aaai.v34i01.5393. Dou, Y.; Shu, K.; Xia, C.; Yu, P.S.; and Sun, L

doi:10.1609/aaai.v34i01.5393

[7] [13]

In Proceedings of the 44th International ACM SIGIR Conference on Research and Develop- ment in Information Retrieval, 2051-2055

User Pref- erence-Aware Fake News Detection. In Proceedings of the 44th International ACM SIGIR Conference on Research and Develop- ment in Information Retrieval, 2051-2055. doi.org/10.1145/3404835.3462990. Su, X.; Yang, J.; Wu, J.; and Zhang, Y

arXiv

[8] [14]

In Proceedings of the sixteenth ACM interna- tional conference on web search and data mining , 51-59

Mining User-Aware Multi-Relations for Fake News Detection in Large Scale Online Social Networks. In Proceedings of the sixteenth ACM interna- tional conference on web search and data mining , 51-59. doi.org/10.1145/3539597.3570478. Zhang, J.; Chen, Y.; Xiao, X.; Lu, R.; and Xia, S

arXiv