Identification of influential spreaders in complex networks using HybridRank algorithm

Area of Science:

• Network Science
• Computational Social Science

Background:

• Identifying influential spreaders is critical for understanding information diffusion in complex networks.
• Applications range from viral marketing to mitigating the spread of harmful content like viruses and cyberbullying.
• Current centrality measures (degree, closeness, betweenness) have limitations in accurately detecting influential spreaders.

Purpose of the Study:

• To propose a novel algorithm, HybridRank, for identifying influential spreaders in complex networks.
• To introduce a new hybrid centrality measure that leverages network topological features.
• To evaluate the effectiveness of HybridRank in detecting influential spreaders.

Main Methods:

• Development of the HybridRank algorithm incorporating a novel hybrid centrality measure.
• Simulation of spreading processes using the SIR (Susceptible-Infected-Recovered) model.
• Empirical evaluation on both real-world and artificial network datasets.

Main Results:

• HybridRank successfully identifies influential spreaders within complex networks.
• The identified spreaders demonstrate greater influence compared to those found using benchmark methods.
• The algorithm's performance is validated through simulations and empirical experiments.

Conclusions:

• The proposed HybridRank algorithm offers a more effective approach to identifying influential spreaders.
• Leveraging hybrid centrality measures enhances the accuracy of influence detection in networks.
• This method has significant implications for optimizing information propagation strategies.