Identifying the influential spreaders in social networks is crucial for maximizing the reach and impact of information dissemination. Many works have been proposed based on undirected and unweighted networks to identify the influential spreaders in the literature. However, many real applications are represented as weighted and directed networks, where the essential characteristics need to be defined for the system. A few works have been proposed based on weighted directed networks, and these measures mainly focused on the concept of finding the central nodes, as reported in the literature. To maximize the spreading propagation in the social network, we have proposed a new “Weighted Cross-bred” method (wcm), which is designed based on the spreading properties of the network. To evaluate the effectiveness of wcm, we have used the Weighted Directed Susceptible-Infected-Recovered (WDSIR) epidemic model as a benchmark simulator, twelve real-world networks, and fourteen well-known existing indexing methods. The results highlight the superior ranking accuracy of wcm, as demonstrated by key performance indicators such as Kendall <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="11227_2025_7658_Article_IEq1.gif" Format="GIF" Height="10" Rendition="HTML" Resolution="72" Type="Linedraw" Width="11" /> </InlineMediaObject> <EquationSource Format="TEX">$\tau$</EquationSource> <EquationSource Format="MATHML"><math> <mi>τ</mi> </math></EquationSource> </InlineEquation> correlation, Precision@r for spreading efficiency, fastest influencer, the effect of various infection probability for seed node-set spreading performance, and average shortest path length <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="11227_2025_7658_Article_IEq2.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="31" /> </InlineMediaObject> <EquationSource Format="TEX">$(L_\mathrm{{s}})$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo stretchy="false">(</mo> <msub> <mi>L</mi> <mi mathvariant="normal">s</mi> </msub> <mo stretchy="false">)</mo> </mrow> </math></EquationSource> </InlineEquation> across different methods. The experimental findings show that the wcm method consistently outperforms other indexing methods regarding spreading dynamics.

Identifying influential spreaders on weighted directed networks based on spreading properties

期刊论文
发表日期：

2025年7月21日

查看全文

Nilanjana Saha,
Amrita Namtirtha,
Animesh Dutta

摘要

Identifying the influential spreaders in social networks is crucial for maximizing the reach and impact of information dissemination. Many works have been proposed based on undirected and unweighted networks to identify the influential spreaders in the literature. However, many real applications are represented as weighted and directed networks, where the essential characteristics need to be defined for the system. A few works have been proposed based on weighted directed networks, and these measures mainly focused on the concept of finding the central nodes, as reported in the literature. To maximize the spreading propagation in the social network, we have proposed a new “Weighted Cross-bred” method (wcm), which is designed based on the spreading properties of the network. To evaluate the effectiveness of wcm, we have used the Weighted Directed Susceptible-Infected-Recovered (WDSIR) epidemic model as a benchmark simulator, twelve real-world networks, and fourteen well-known existing indexing methods. The results highlight the superior ranking accuracy of wcm, as demonstrated by key performance indicators such as Kendall $\tau$ $τ$ correlation, Precision@r for spreading efficiency, fastest influencer, the effect of various infection probability for seed node-set spreading performance, and average shortest path length $(L_\mathrm{{s}})$ $(L_{s})$ across different methods. The experimental findings show that the wcm method consistently outperforms other indexing methods regarding spreading dynamics.