Balancing Speed and Coverage by Sequential Seeding in Complex Networks

Area of Science:

• Network Science
• Information Diffusion Models

Background:

• Information spreading in complex networks is commonly modeled as diffusion from activated nodes to neighbors.
• Information cascades occur when initial seed nodes trigger widespread diffusion.

Purpose of the Study:

• To introduce and evaluate novel sequential seed initiation strategies for information diffusion.
• To compare sequential seeding with traditional single-stage seeding approaches.

Main Methods:

• Developed sequential seeding strategies that activate nodes in stages, avoiding already diffused nodes.
• Compared sequential and single-stage seeding using various ranking methods and diffusion parameters on real networks.

Main Results:

• Sequential seeding strategies outperformed single-stage seeding, achieving better coverage in approximately 90% of tested scenarios.
• Longer seeding sequences increased activated nodes but also extended diffusion time.
• Different sequential variants offered varied trade-offs between coverage and diffusion speed.

Conclusions:

• Sequential seeding is a more effective approach for maximizing information diffusion coverage in complex networks.
• The timing of seed activation significantly impacts the efficiency and outcome of information cascades.
• Optimizing sequential seeding strategies can balance network coverage with diffusion speed.