🤖 AI Summary
Existing models of complex contagion struggle to capture the interplay among individual preferences, local social influence, and global sentiment, and offer limited insight into the critical thresholds governing phase transitions in viral spread. This work proposes a unified cascade model that embeds both ideas and network nodes into a shared high-dimensional feature space. Node state updates are driven by a decision function integrating transmission affinity, local reinforcement, and global activation, yielding an efficiently samplable Markovian cascade process. The model reveals, for the first time, how the dynamic balance between local and global influences critically determines cascade success or failure, and demonstrates that early-stage growth patterns can effectively predict phase transitions. Comprehensive experiments analyze cascade distributions, latent dynamics, parameter sensitivity, and critical behavior, establishing a new paradigm for studying complex contagion mechanisms.
📝 Abstract
Understanding how complex behaviors, opinions, and innovations spread in online social networks remains a central challenge in computational social science. Existing models of complex contagion typically rely on stylized threshold mechanisms based solely on the number of infected neighbors and do not account for the interaction between individual preferences, local social influence, and global sentiment. Moreover, the emergence of virality through phase transitions and tipping points remains poorly characterized.
In this paper, we propose a unified propagation cascade model in which notions propagate as high-dimensional vectors in the same feature space as network nodes. Node activations are governed by a unified decision function that integrates propagation affinity, local influence, and global influence. The resulting dynamics induce a stochastic, Markovian cascade process that enables efficient MCMC sampling of propagation outcomes.
Using preferential attachment networks, we systematically study spread distributions, incubation dynamics, parameter sensitivity, and phase transition behavior. Our results show that balanced interactions between local reinforcement and global activation are critical for successful cascades and that early-stage growth patterns provide reliable signals of impending phase transitions.