Source inference for misinformation spreading on hypergraphs DOI
Xiaohang Yu, Yanyi Nie, Wenyao Li

и другие.

Chaos Solitons & Fractals, Год журнала: 2024, Номер 187, С. 115457 - 115457

Опубликована: Авг. 31, 2024

Язык: Английский

Network alignment DOI
Rui Tang, Ziyun Yong, Shuyu Jiang

и другие.

Physics Reports, Год журнала: 2024, Номер 1107, С. 1 - 45

Опубликована: Дек. 2, 2024

Процитировано

3

Modelling multiscale infectious disease in complex systems DOI
Jiajun Xian, Minghui Liu,

Xuan Cheng

и другие.

Physics Reports, Год журнала: 2025, Номер 1113, С. 1 - 57

Опубликована: Фев. 11, 2025

Язык: Английский

Процитировано

0

Integrating Message Content and Propagation Path for Enhanced False Information Detection Using Bidirectional Graph Convolutional Neural Networks DOI Creative Commons
Jie Hu, Gang Yang, Binbin Tang

и другие.

Applied Sciences, Год журнала: 2025, Номер 15(7), С. 3457 - 3457

Опубликована: Март 21, 2025

We investigate the impact of textual content and its structural characteristics on detection false information. propose a Bidirectional Graph Convolutional Neural Network (ICP-BGCN) that integrates message with propagation paths for enhanced performance. Our approach leverages web topology by transforming disconnected user posts into bidirectional graph, which top-down bottom-up pathways derived from post forwarding commenting relationships. Using BERT embeddings, we extract contextual semantic features both source texts their propagated counterparts, are embedded as node attributes within graph. The graph convolutional neural network subsequently learns feature representations event during information dissemination, merging these original text to achieve comprehensive disinformation detection. Experimental results demonstrate significant improvements over existing methods. On benchmark datasets Twitter15 Twitter16, our model achieves accuracy rates 89.7% 91.7%, respectively, outperforming state-of-the-art baselines 1.1% 3.7%. proposed ICP-BGCN exhibits strong cross-domain generalization, attaining 84.4% Pheme dataset achieving 1.8% in 3.8% Macro-F1 score SemEval-2017 Task 8.

Язык: Английский

Процитировано

0

Heterogeneous K-core percolation on hypergraphs DOI
Dandan Zhao,

W. Xi,

Bo Zhang

и другие.

Chaos An Interdisciplinary Journal of Nonlinear Science, Год журнала: 2025, Номер 35(3)

Опубликована: Март 1, 2025

In complex systems, there are pairwise and multiple interactions among elements, which can be described as hypergraphs. K-core percolation is widely utilized in the investigation of robustness systems subject to random or targeted attacks. However, nodes usually correlates with their characteristics, such degree, exhibits heterogeneity while lacking a theoretical study on hypergraph. To this end, we constructed hyperedge model that introduces parameters divide active hyperedges into two parts, where inactive unless they have certain number nodes. stage pruning process, when contained less than its set value, it will pruned, result deletion other ultimately trigger cascading failures. We studied magnitude giant connected component threshold by mapping hypergraph factor graph. Subsequently, conducted large simulation experiments, values matched well simulated values. The proposed significant impact type phase transition network. found decreasing value renders network more fragile, increasing makes resilient under Meanwhile, parameter decreases 0, may cause change nature transition, shows hybrid transition.

Язык: Английский

Процитировано

0

Modeling two competing infectious diseases in a metropolitan contact network DOI
Yue Xiao, Wenjie Li, Yanyi Nie

и другие.

Chaos Solitons & Fractals, Год журнала: 2025, Номер 196, С. 116282 - 116282

Опубликована: Апрель 7, 2025

Язык: Английский

Процитировано

0

Bootstrap percolation on hypergraph DOI
Hao Peng,

Chenglong Wang,

Dandan Zhao

и другие.

Chaos An Interdisciplinary Journal of Nonlinear Science, Год журнала: 2025, Номер 35(4)

Опубликована: Апрель 1, 2025

Bootstrap percolation is a widely studied model to investigate the robustness of network for cascading failures. Extensive real-world data analysis has revealed existence higher-order interactions among elements, i.e., beyond pairwise, which are usually described by hypergraphs. In this paper, we propose generalized bootstrap on hypergraph, assumes that activation an inactive node depends number active neighbors through its hyperedges. Through numerical simulation and theoretical analysis, found threshold phase transition type closely related infection proportion edges. When significant, any initial probability, size giant component (GAC) shows continuous growth with increasing occupation probability. small, increase GAC changes from discontinuous growth. addition, in case fixed average degree, edges will reduce threshold, conducive enhancing network. At same time, create more opportunities nodes be activated, under conditions change

Язык: Английский

Процитировано

0

Learning influence probabilities in diffusion networks without timestamps DOI
Yuchen Wang, Huidi Wang, Chao Gao

и другие.

Applied Mathematics and Computation, Год журнала: 2025, Номер 503, С. 129502 - 129502

Опубликована: Май 8, 2025

Язык: Английский

Процитировано

0

Robustness of multilayer interdependent higher-order network DOI
Hao Peng, Yifan Zhao, Dandan Zhao

и другие.

Journal of Network and Computer Applications, Год журнала: 2024, Номер unknown, С. 104047 - 104047

Опубликована: Окт. 1, 2024

Язык: Английский

Процитировано

2

Source inference for misinformation spreading on hypergraphs DOI
Xiaohang Yu, Yanyi Nie, Wenyao Li

и другие.

Chaos Solitons & Fractals, Год журнала: 2024, Номер 187, С. 115457 - 115457

Опубликована: Авг. 31, 2024

Язык: Английский

Процитировано

1