[Paper Review] Epidemic Dynamics On Information-Driven Adaptive Networks

Xiu‐Xiu Zhan, Chuang Liu+2 more|arXiv (Cornell University)|May 19, 2015

Complex Network Analysis Techniques32 references3 citations

TL;DR

This paper proposes an information-driven adaptive network model where susceptible individuals sever links with infected neighbors upon learning of disease risks, combining epidemic spreading (SI/SIS) with information diffusion. Simulations and pairwise modeling show this adaptive behavior significantly reduces epidemic prevalence and delays spreading, with bifurcation analysis revealing complex dynamics across healthy, oscillatory, bistable, and endemic regions.

ABSTRACT

can evolve simultaneously. For the information-driven adaptive process, susceptible (infected) individuals who have abilities to recognize the disease would break the links of their infected (susceptible) neighbors to prevent the epidemic from further spreading. Simulation results and numerical analyses based on the pairwise approach indicate that the information-driven adaptive process can not only slow down the speed of epidemic spreading, but can also diminish the epidemic prevalence at the final state significantly. In addition, the disease spreading and information diffusion pattern on the lattice give a visual representation about how the disease is trapped into an isolated field with the information-driven adaptive process. Furthermore, we perform the local bifurcation analysis on four types of dynamical regions, including healthy, oscillatory, bistable and endemic, to understand the evolution of the observed dynamical behaviors. This work may shed some lights on understanding how information affects human activities on responding to epidemic spreading.

Motivation & Objective

• To model the interplay between epidemic spreading and disease information diffusion on adaptive networks where individuals dynamically cut links with infected neighbors.
• To investigate how information-driven adaptive rewiring influences epidemic dynamics beyond static network assumptions.
• To analyze the emergence of complex dynamical behaviors—such as oscillations and bistability—arising from the feedback between information diffusion and network adaptation.
• To quantify the impact of information on epidemic thresholds and final prevalence using both simulation and analytical pairwise approaches.

Proposed method

• Proposes a four-state model: S− (susceptible, unaware), S+ (susceptible, aware), I− (infected, unaware), I+ (infected, aware), with transitions driven by infection, awareness, and adaptive link-breaking.
• Introduces adaptive link-cutting: S− individuals break links with I− neighbors at rate α, while I− individuals break links with S− at rate ω, based on awareness of infection.
• Employs a pairwise mean-field approximation to derive a system of 13 coupled ODEs tracking the time evolution of node pairs (e.g., [S−I−], [S+I+]), incorporating transmission, awareness, and rewiring rates.
• Uses local bifurcation analysis to classify dynamical regions (healthy, oscillatory, bistable, endemic) based on parameters like infection rate β, awareness rate σ, and link-cutting rates α, ω.
• Performs agent-based simulations on regular lattices to validate analytical results and visualize spatial patterns of isolated disease clusters.
• Analyzes the mutual feedback loop: infected individuals generate information, which triggers adaptive behavior, reducing SI interactions and altering epidemic outcomes.

Experimental results

Research questions

• RQ1 How does information-driven adaptive link-breaking affect the final epidemic prevalence and spreading speed?
• RQ2 What dynamical regimes (e.g., oscillations, bistability) emerge in the epidemic dynamics due to the feedback between awareness and network adaptation?
• RQ3 How does the adaptive process alter the epidemic threshold and the stability of disease-free and endemic states?
• RQ4 In what ways does the spatial structure of the network (e.g., lattice) influence the formation of isolated disease clusters under adaptive behavior?
• RQ5 What is the relative contribution of reduced transmission versus link-cutting in suppressing epidemic spread?

Key findings

• The information-driven adaptive process significantly reduces the final epidemic prevalence by isolating infected individuals through link-breaking behavior.
• Simulations show that disease spreading is slowed and eventually trapped in isolated clusters due to the formation of 'information barriers' that prevent further transmission.
• The pairwise model accurately predicts the epidemic threshold and final prevalence, with bifurcation analysis identifying four distinct dynamical regions: healthy, oscillatory, bistable, and endemic.
• The model reveals that link-cutting (α, ω) plays a crucial role in suppressing epidemics, with higher rates leading to earlier disease extinction and lower final prevalence.
• The coexistence of multiple stable states (bistability) and oscillatory behavior emerges only in adaptive networks, not in static networks, highlighting the role of network dynamics in shaping epidemic outcomes.
• Numerical results confirm that both reduced transmission (via awareness) and active link-cutting contribute synergistically to epidemic control, with the latter being particularly effective in limiting SI contact formation.