Two-step seeding strategy in multiplex networks with inter-layer conversion cost of influence.

Area of Science:

• Network Science
• Information Diffusion
• Computational Social Science

Background:

• Multilayer networks present complex dynamics for influence propagation.
• Inter-layer conversion costs complicate traditional seeding strategies.
• Identifying key individuals in multiplex networks is crucial for effective information dissemination.

Purpose of the Study:

• To propose a novel two-step influence seeding strategy for multiplex networks.
• To address the challenge of inter-layer conversion costs in influence diffusion.
• To enhance influence coverage by accurately identifying influential seeds.

Main Methods:

• Developed a two-step strategy: target layer selection and target seed identification.
• Proposed multiplex betweenness centrality and multi-hop multiplex neighbors (MMNs) for layer selection.
• Introduced multiplex gravity centrality (MGC) for seed identification within the target layer.
• Utilized a redefined independent cascade model for evaluating seeding strategy effectiveness.

Main Results:

• The proposed seeding strategy significantly increases influence coverage.
• Experimental validation on synthetic and real-world datasets confirms the strategy's effectiveness.
• MMN-based layer selection demonstrates robustness across different neighborhood ranges.
• MGC combined with MMNs achieves superior influence performance with optimal parameter settings.

Conclusions:

• The novel two-step seeding strategy is effective for maximizing influence in multiplex networks.
• The proposed multiplex gravity centrality offers a powerful tool for identifying influential nodes.
• The findings provide valuable insights for optimizing information diffusion strategies in complex network structures.