Source-linked AI summary
Suppressing disease spreading by using information diffusion on multiplex networks
Wei Wang, Quan-Hui Liu, Shi-Min Cai, Ming Tang, Lidia A. Braunstein, H. Eugene Stanley
TL;DR
The paper addresses the limited empirical understanding of how information and disease spreading coevolve. Using real influenza-related data and an asymmetric multiplex-network model, it finds that disease can trigger information outbreaks, while information has an optimal transmission rate that markedly suppresses disease without changing its threshold.
Problem
The paper addresses the lack of systematic empirical understanding of the microscopic mechanisms connecting information and disease spreading.
Method
It combines real-world influenza-like-illness data with an asymmetric information–disease spreading model on multiplex networks, analyzed using heterogeneous mean-field theory and stochastic simulations.
Results
An optimal information transmission rate markedly suppresses disease spreading, while information spreading does not alter the disease threshold.
Takeaways & Limitations
The results quantify how information diffusion can be used to suppress disease while explaining asymmetric coevolution and an unchanged disease threshold.
Takeaways & Limitations
The theoretical treatment is limited by strong dynamic correlations among neighboring node states, motivating more accurate methods such as dynamic message passing or pair approximation.
Abstract
from arXiv · showhide
Although there is always an interplay between the dynamics of information diffusion and disease spreading, the empirical research on the systemic coevolution mechanisms connecting these two spreading dynamics is still lacking. Here we investigate the coevolution mechanisms and dynamics between information and disease spreading by utilizing real data and a proposed spreading model on multiplex network. Our empirical analysis finds asymmetrical interactions between the information and disease spreading dynamics. Our results obtained from both the theoretical framework and extensive stochastic numerical simulations suggest that an information outbreak can be triggered in a communication network by its own spreading dynamics or by a disease outbreak on a contact network, but that the disease threshold is not affected by information spreading. Our key finding is that there is an optimal information transmission rate that markedly suppresses the disease spreading. We find that the time evolution of the dynamics in the proposed model qualitatively agrees with the real-world spreading processes at the optimal information transmission rate.
RESULTS
Real-data analysis and multiplex-network modeling reveal asymmetric coevolution: disease can promote information spreading, while information-based immunization suppresses disease without changing its outbreak threshold. The model identifies an optimal information transmission rate that markedly reduces disease spreading and reproduces observed coevolution patterns.
- Real-world agreement: At the optimum, model time evolution qualitatively agrees with real-world data, showing both same-direction and opposite-direction short-window growth correlations.The simulated cross-correlations reproduce the coexistence of positive and negative growth trends observed in the influenza-like-illness data.
- Disease threshold: The disease threshold remains λB_c = 1/⟨kB⟩, because vaccination requires authentication from both layers and is negligible below the disease threshold.For φ ≥ 1, informed counterparts cannot be immunized until their contact-layer infection conditions are also satisfied.
- Disease threshold: The disease threshold is unchanged by information-based immunization and depends only on the contact-layer topology, not communication-layer topology, p, or φ.This matches the SIR threshold without immunization and differs from earlier results.
- Theory and simulation: The heterogeneous mean-field predictions agree with stochastic simulations, with residual differences attributed to dynamic neighbor-state correlations and finite network size.These correlations arise when transmission events to one node from distinct neighbors are correlated.
- Asymmetric interactions: Information outbreaks can arise from information spreading itself or from disease outbreaks, while increasing disease transmission increases information spreading.For λB = 0.2, 0.5, and 0.8, any λA can trigger information spreading because disease has already spread.
- Optimal suppression: An optimal information transmission rate λA maximizes vaccination and minimizes final disease size RB.The optimum arises because faster information first increases informed counterparts but, when too rapid, prevents the infection-neighbor threshold from being met.
DISCUSSION
The study combines real-data evidence with multiplex-network modeling to characterize asymmetric information–disease coevolution and identify mechanisms for suppressing disease spread.
- Information outbreaks can arise from information spreading itself or from disease outbreaks, whereas information spreading does not affect the disease threshold.
- The data-driven asymmetric mechanism supports earlier modeling assumptions and reveals coevolution phenomena that differ qualitatively from independent-spreading disease-behavior systems.
- Further research should incorporate additional real-world mechanisms and stronger methods for dynamic correlations among neighboring network nodes.The authors identify dynamic message passing and pair approximation as possible approaches.
METHODS
The methods define relative growth rates for the information and disease time series.
- Relative growth rates are defined for the information series nG(t) and disease series nD(t).
FIGURE LEGENDS
The figure tracks normalized information and disease activity, their growth rates, cross-correlations, and correlation fractions across time windows.
- The figure compares normalized outpatient visits and Google Flu Trends search queries over time, alongside their relative growth rates.
- Cross-correlation c(t) is plotted for window sizes wl = 3 and wl = 20, while correlation fractions vary with wl.
- The figure marks relative growth rates at t = 53 and t = 153 using circles and squares, respectively.
- Additional parameter settings include φ = 2 and p = 0.8.