Abstract:In recent years, studies in many fields have widely explored diffusion in social networks, which refers to the collective behavior of a set of autonomous social actors who interact with something (such as opinions, viruses, or innovations) in social networks. Many researchers have utilized a multiagent method to explore the effects of social actor characteristics and behavioral complexity on information diffusion. This method performs well in modeling the complex characteristics and decision-making of social actors. Although this subject has been intensively studied, analyses have typically been limited to concerns about the effects of entire social network structures, such as the small-world and scale-free aspects. Few empirical information diffusion studies have systematically investigated causal effects and decision-making mechanisms under the constraints of the interaction structures of agents. In practice, we observed that the decision-making of each agent in information diffusion was significantly influenced and constrained by its contextual network structures. To solve this problem, this study utilizes the cause–effect graph method to model the interaction structures of agents in social networks and presents agent decision-making models under the constraints of local network structures in information diffusion. The proposed models and methods can be used to effectively analyze the mechanism of the influence of structured interaction relations of agents on the process of decision-making and diffusion results. Moreover, the presented decision-making model can comprehensively consider the factors of social positions of heterogeneous agents, interaction strengths, and diffusion strategies, and thus, can better match the features of real-world influence diffusion in social networks.

Modeling and maximizing information diffusion over hypergraphs based on deep reinforcement learning

Non-Progressive Influence Maximization in Dynamic Social Networks

Deep Graph Representation Learning and Optimization for Influence Maximization

Deep graph representation learning influence maximization with accelerated inference

Identifying Critical Nodes Via Link Equations and Deep Reinforcement Learning

H-Diffu: Hyperbolic Representations for Information Diffusion Prediction

Finding Influencers in Complex Networks: An Effective Deep Reinforcement Learning Approach

DSCom: A Data-Driven Self-Adaptive Community-Based Framework for Influence Maximization in Social Networks

Modeling information diffusion dynamics over social networks

Information Spreading Dynamics on Adaptive Social Networks

Maximizing Social Influence with Minimum Information Alteration

Hierarchical Community-Level Information Diffusion Modeling in Social Networks

DISCO: Influence Maximization Meets Network Embedding and Deep Learning

Multi-task Learning for Influence Estimation and Maximization

Evolutionary social information diffusion analysis

Balanced influence maximization in social networks based on deep reinforcement learning

Structured multiagent decision-making in information diffusion: The model and dynamics

Full-Scale Information Diffusion Prediction With Reinforced Recurrent Networks

Evolutionary Dynamics of Information Diffusion over Social Networks

Modeling Multi-source Information Diffusion: A Graphical Evolutionary Game Approach

Evolutionary Information Diffusion over Social Networks