Product diffusion in dynamic online social networks: A multi-agent simulation based on gravity theory

Abstract

The popularity of mobile social media has promoted the process of consumers accepting new products online, that is, online new product diffusion (ONPD). However, previous research has seldom considered the dynamic social networks (varying-weight) resulting from the gravitational field generated by consumer interactions. To explore the effect of varying-weight social networks on ONPD, we propose a Brownian agent simulation model integrating the gravitational field theory and evolutionary game. Specifically, the evolutionary game theory is utilized to describe the interactions between consumers. Then the gravitational field theory is introduced to investigate the gravitational field generated by consumer interactions, which serves as a formation mechanism of a varying-weight social network. Then, a Brownian agent ONPD model is realized under different network structures. The effectiveness of the simulation model proposed is validated by comparison with real-world cases. The results show that: (1) The gravitational coefficient (represents the attraction of members in an online community) positively affects the diffusion outcome in the social network, but a moderate gravitational coefficient can accelerate the diffusion in the early stage. (2) Hub nodes can effectively moderate the negative effect of consumption inertia on product diffusion. (3) The network clustering coefficient can enhance the positive effect of the trust coefficient on product diffusion. (4) The negative impact of information noise (e.g., fake news, false information in online social media) on product diffusion is aggravated by network heterogeneity. This integrated study provides a new perspective to explore ONPD by applying the gravitational field theory to the diffusion model.