Research on Social Marketing Strategies with An Agent-based Propagation Model

Considering the increasing complexity of social networks and user behaviors, it is very challenging for advertisers to formulate their strategies of selecting proper initial seed users in their social marketing efforts. In this paper, we tackle this challenge by proposing an agent-based propagation model and injecting it into typical social networks with three types of structures, i.e., Erdos-Renyi random graph, Watt-Strogatz small world graph, and Barabasi-Albert scale-free graph. We instantiate this agent-based model with demographic characteristics extracted from real-world census data collected in China. By investigating the diffusion process of advertising infounation in these social networks, we can analyze and compare the perfounance of advertisers' targeting and influencer strategies. Our experimental results indicate that advertisers adopting influencer strategies should manipulate the initial well-connected seeds to deliver infounation only to the potential customers instead of a wide range of generic users. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.