GDLC: A new Graph Deep Learning framework based on centrality measures for intrusion detection in IoT networks

Internet of Things, Год журнала: 2024, Номер 26, С. 101214 - 101214

Опубликована: Май 7, 2024

Язык: Английский

---

A multilevel backbone extraction framework

Applied Network Science, Год журнала: 2024, Номер 9(1)

Опубликована: Июль 22, 2024

Abstract As networks grow in size and complexity, backbones become an essential network representation. Indeed, they provide a simplified yet informative overview of the underlying organization by retaining most significant structurally influential connections within network. Network heterogeneity often results complex intricate structures, making it challenging to identify backbone. In response, we introduce Multilevel Backbone Extraction Framework, novel approach that diverges from conventional backbone methodologies. This generic prioritizes mesoscopic networks. First, splits into homogeneous-density components. Second, extracts independent for each component using any classical technique. Finally, various are combined. strategy effectively addresses observed groupings. Empirical investigations on real-world underscore efficacy preserving structures properties. Experiments demonstrate its superiority over methods handling enhancing integrity. The framework is adaptable types extraction techniques, versatile tool analysis across diverse applications.

Язык: Английский

---

Identifying influential airports in airline network based on failure risk factors with TOPSIS

Chaos Solitons & Fractals, Год журнала: 2023, Номер 169, С. 113310 - 113310

Опубликована: Март 6, 2023

Язык: Английский

---

Transformations in the European Gas Supply Network Due to the Russia–Ukraine Conflict

Energies, Год журнала: 2025, Номер 18(7), С. 1709 - 1709

Опубликована: Март 28, 2025

The objective of this paper is to demonstrate the structural characteristics European gas supply system and changes in its network structure interaction clustering among nodes defined as countries, following outbreak Russia–Ukraine conflict. methodology relies on social analysis, which employs mathematics graph theory examine state dynamics given structure. impacts identified involve reduced reliance Russian gas, a considerable reduction strength centrality Russia Germany, higher dispersion flows, largely due increased import LNG flows. After conflict outbreak, countries such Italy, Austria, Slovak Republic, Hungary retained their high influential position, terms PageRank centrality, while Balkan together with Middle East suppliers (Turkey Iran), formed common group Russia. estimated stress challenges posed EU enhance connectivity infrastructure investments explore alternative ways support objectives strategic autonomy, promoting resilience path toward carbon-free transition.

Язык: Английский

---

Influential nodes identification for complex networks based on multi-feature fusion

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 3, 2025

Язык: Английский

---

Component Structure-Driven Centrality Ranking for Diffusion Management in Complex Networks

Studies in computational intelligence, Год журнала: 2025, Номер unknown, С. 53 - 65

Опубликована: Янв. 1, 2025

Язык: Английский

---

A comparative study of methods for measuring node influence in complex networks

Engineering Applications of Artificial Intelligence, Год журнала: 2025, Номер 155, С. 111088 - 111088

Опубликована: Май 12, 2025

Язык: Английский

---

CGNLib: A Python library for Girvan–Newman community detection with customizable node-based centrality metrics

SoftwareX, Год журнала: 2025, Номер 31, С. 102193 - 102193

Опубликована: Май 24, 2025

Язык: Английский

---

Improving the controllability robustness of complex temporal networks against intelligent attacks

Journal of Complex Networks, Год журнала: 2024, Номер 12(4)

Опубликована: Июнь 24, 2024

Abstract The main goal of controllability network methods on complex temporal networks is to control all nodes with the minimum number nodes. Real-world are faced many errors and attacks that cause structure be changed in some way so processes disturbed after that, robustness decreases. One most important intelligent attacks. In this paper, types their destructive effects have been investigated. order increase against attacks, a novel graph model strategies proposed dynamic by adding new or links protected results simulation comparing them conventional demonstrate node addition strategy has performed better than other improvement rate terms execution time 60%. On hand, immunization kept controllable smaller (38%) less (52%) compared methods.

Язык: Английский

---

Convex Isolating Clustering Centrality to Discover the Influential Nodes in Large Scale Networks

IEEE Access, Год журнала: 2024, Номер 12, С. 48404 - 48419

Опубликована: Янв. 1, 2024

Ranking influential nodes within complex networks offers invaluable insights into a wide array of phenomena ranging from disease management to information dissemination and optimal routing in real-time networking applications. Centrality measures, which quantify the importance based on network properties relationships network, are instrumental achieving this task. These measures typically classified local global centralities. Global consider overall structure connectivity patterns. However, they often suffer high computational complexity large-scale networks. On other hand, focus immediate neighborhood each node, potentially overlooking information. To address these challenges, we propose novel metric called Isolating Clustering (ISCL), leverages convex combination approach. By introducing tuning parameter, ISCL enhances applicability adaptability centrality across range real-world In study, assess efficacy proposed measure using datasets simulate spreading process susceptible-infected-removed (SIR) independent cascade (IC) models. Our extensive results demonstrate that significantly improves efficiency compared conventional recent while also maintaining better

Язык: Английский

---