SpreadRank: A Novel Approach for Identifying Influential Spreaders in Complex Networks

Area of Science:

• Network Science
• Information Diffusion Modeling

Background:

• Identifying influential spreaders is vital for managing information spread and suppressing malware.
• Existing methods may not fully capture node influence or account for overlapping impact areas.

Purpose of the Study:

• To propose a novel method, SpreadRank, for identifying influential spreaders in complex networks.
• To enhance the selection of influential nodes by considering path reachability and minimizing influence overlap.

Main Methods:

• SpreadRank quantifies node spread influence using path reachability and spread centrality.
• A dynamic node set selection mechanism is employed, minimizing maximum connected branches after network segmentation.
• Node influence is evaluated based on spread centrality and network partitioning.

Main Results:

• SpreadRank effectively identifies influential spreaders for information diffusion and suppression.
• Experimental results using the SIR model show superior performance compared to existing methods.
• The selected SpreadRank nodes demonstrate enhanced ability to rapidly spread or control information.

Conclusions:

• SpreadRank offers a more effective approach to identifying influential spreaders in complex networks.
• The method's consideration of path reachability and dynamic selection improves overall spread influence management.
• SpreadRank provides a valuable tool for applications requiring targeted information dissemination or containment.