Powerful Influencer Identification in Temporal Social Networks Adopting Discrete Wild Geese Swarm Optimization

Influencer Identification has become a crucial fragment for nowadays online product promotion. Traditional meta-heuristic solutions can bring a stable trade-off between performance and efficiency. However, most of them only focus on static application scenarios. To tackle this issue, in this paper, we propose a new algorithm for powerful influencer identification in temporal social networks. Firstly, we delicately design a metric named Stability Centrality (SC) to generate the potential candidate seeds. Then, we formally define a novel fitness function called Temporal Local Influence Estimation (TLIE) to accurately and rapidly evaluate the quality of the candidate solutions. Afterwards, enlightened by the biological wild geese queue effect in nature, based on the SC metric and the TLIE function, we present an innovative algorithm named Discrete Wild Geese Swarm Optimization (DWGSO) for powerful spreader recognition in dynamic temporal networks. Experimental results on four real-world temporal networks as well as the ablation experiments have demonstrated the capacity of our proposed DWGSO for effectively and efficiently identifying persuasive disseminator in temporal networks.