Influential Node Detection Based on Implicit Communities
Neda Binesh, Mehdi Ghatee · 2025
Finding influential nodes in social networks is crucial for spreading information. This chapter focuses on the linear threshold (LT) model, which suggests that individuals are more likely to be influenced when many of their neighbors are active. However, if all the influencers come from one community, they may struggle to spread their influence beyond their community. The challenge here lies in the vast size of social networks and in identifying influencers who can activate people across different communities. To address these challenges, we explore the identification of implicit communities within a network containing candidate nodes based on their social distance. In addition, to maximize influence in the LT model, we propose a mathematical model for distributing influencers across communities and selecting those with many common neighbors within their community. This includes defining and computing a new social distance for candidate nodes and using a Laplacian matrix to identify implicit communities and select the best influencers within them. We also outline the theorems that can help discover the size of communities in a candidate node network based on distance information.