Infection patterns in simple and complex contagion processes on networks.

Area of Science:

• Complex Systems Science
• Network Science
• Epidemiology
• Computational Social Science

Background:

• Spreading processes, such as disease or information diffusion, are commonly studied on interaction networks.
• While network structure's influence on spread is well-researched, the converse—how different contagion processes affect infection patterns on a fixed network—remains less explored.

Purpose of the Study:

• To investigate how various contagion models and their parameters influence infection patterns on a given network.
• To understand the relationship between contagion process characteristics and emergent spreading dynamics.

Main Methods:

• Simulation of diverse contagion models (simple contagion, complex contagion, threshold mechanisms) on defined network structures.
• Analysis of infection patterns generated by each model, focusing on parameter dependencies and variations in spreading paths.

Main Results:

• Simple contagion processes exhibit highly robust infection patterns, largely independent of model parameters.
• Complex contagion models demonstrate non-trivial dependencies, with infection patterns influenced by the balance of pairwise and group contagions.
• Threshold-based models show significant sensitivity, where minor parameter changes can drastically alter spreading pathways.

Conclusions:

• Infection patterns are not solely determined by network structure but are significantly shaped by the nature of the contagion process.
• Schematized models can reveal crucial features of spread, but understanding variations requires considering model-specific parameters and contagion types.
• Divergent spreading patterns arise from different contagion mechanisms, underscoring the importance of model choice in studying diffusion phenomena.