Misinformation Dynamics in Social Networks

TL;DR

This work addresses how information fidelity degrades as it propagates through multiplex social networks comprising private chats, group chats, and broadcast channels. It introduces a continuous fidelity field $F_i(t) \in [0,1]$ on a three-layer network and a governing dynamics $\frac{dF_i}{dt} = -\delta F_i - \beta F_i^2 + \sum_{\ell=1}^{3} \Gamma_{\ell} D_i^{(\ell)}[\{F_j\}]$, with a nonlinear groupthink term $D^{(2)}$. The paper identifies three universal mechanisms—groupthink blending, bridge-node bottlenecks, and a network-wide fidelity landscape—and derives a closed-form steady-state fidelity $\langle F^* \rangle = \frac{D_3 p_b}{D_3 p_b + \delta + \xi \langle k_{bridge} \rangle f(\langle m\rangle)}$, alongside a mean-field phase diagram mapping regimes in $(\langle m\rangle, D_3)$. These results provide a quantitative framework linking network topology to information integrity and offer design principles for mitigating misinformation in large-scale communication platforms.

Abstract

Information transmitted across modern communication platforms is degraded not only by intentional manipulation (disinformation) but also by intrinsic cognitive decay and topology-dependent social averaging (misinformation). We develop a continuous-fidelity field theory on multiplex networks with distinct layers representing private chats, group interactions, and broadcast channels. Our analytic solutions reveal three universal mechanisms controlling information quality: (i) groupthink blending, where dense group coupling drives fidelity to the initial group mean; (ii) bridge-node bottlenecks, where cross-community flow produces irreversible dilution; and (iii) a network-wide fidelity landscape set by a competition between broadcast truth-injection and structural degradation pathways. These results demonstrate that connectivity can reduce information integrity and establish quantitative control strategies to enhance fidelity in large-scale communication systems.

Figures (4)

• Figure 1: Multiplex topology of a WhatsApp-like network. Nodes represent users and edges represent three channel types: private 1-to-1 chats (thin undirected edges), group cliques (dense patches), and directed broadcast connections (arrows). Black-rimmed nodes are broadcasters. This structure defines the heterogeneous coupling operators in Eq. \ref{['fid_dyn']}.
• Figure 2: Characteristic fidelity decay curve. Rapid initial drop reflects fast group-blending dynamics, governed by $mD_2$; slow recovery arises from broadcast-driven reinforcement $D_3$.
• Figure 3: Steady-state average fidelity $\langle F^*\rangle$ is shown as a function of broadcast coverage $p_b$, average group size $\langle m\rangle$, and average bridge connectivity $\langle k_{\mathrm{bridge}}\rangle$. Increased intergroup connectivity and large group size drive fidelity loss.
• Figure 4: Structural phase diagram showing regimes of information fidelity as a function of average group size $\langle m\rangle$ and broadcast strength $D_3$. In the red region, rapid group averaging overwhelms broadcast and private correction ($\tau_g < \tau_b, \tau_g < \tau_p$), leading to a groupthink-driven fidelity collapse. In the blue region, broadcast reinforcement arrives soon enough to prevent collapse. The dashed boundary ($\tau_b=\tau_p$) defines the operational stability threshold from Eq. \ref{['dynamical_regime']}.