STC plus K: a Semi-global triangular and degree centrality method to identify influential spreaders in complex networks

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2023

Autores

Sadhu, S (Autor) Outra; Namtirtha, A (Autor) Outra; Malta, MC (Autor) Outra; Dutta, A (Autor) Outra

Ata de Conferência Internacional

Título: 2023 IEEE INTERNATIONAL CONFERENCE ON WEB INTELLIGENCE AND INTELLIGENT AGENT TECHNOLOGY, WI-IAT

Páginas: 655-662

22nd IEEE/WIC International Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT)

Venice, ITALY, OCT 26-29, 2023

Indexação

ISI Web of Knowledge - 0 Citações

Scopus - 0 Citações

Outras Informações

ID Authenticus: P-00Z-VXM

DOI: 10.1109/wi-iat59888.2023.00107

Abstract (EN): Influential spreaders contribute substantially to managing and optimizing any spreading process in a network. Influential spreaders are nodes that hold importance within the network. Identifying them is a challenging task. Some encysting methods for such identification include local-structure-based, global-structure-based, semi-global-structure-based, and hybrid-structure-based methods. Semi-global structure-based methods show significant potential in identifying influential nodes in different network structures. However, existing semi-global structure-based methods often identify nodes from the network's periphery, where nodes are loosely connected, and their collective influence in spreading processes is minimal. This paper presents a novel method called Semi-global triangular and degree centrality (STC + K) to overcome this limitation by considering a node's degree, the number of triangles, and the third hop of neighbourhood connectivity information. The proposed novel method outperforms the existing noteworthy indexing methods regarding ranking performance. The experimental results show better performance, as indicated by two performance metrics: recognition rate and improvement percentage. By virtue of the fact that the empirically set free parameters are absent, our method eliminates the need for time-consuming preprocessing to select optimal parameter values for ranking nodes in large networks.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Nº de páginas: 8