Large-scale influence maximization via maximal covering location

Influence maximization aims at identifying a limited set of key individuals in (social) network which spreads information based on some propagation model and maximizes the number reached. We show that influence probabilistic independent cascade can be modeled as stochastic maximal covering location problem. A reformulation Benders decomposition is proposed relation between obtained optimality cuts submodular for correspondingly defined subsets established. introduce preprocessing tests, allow us to remove variables from develop efficient algorithms separation cuts. Both aspects are shown crucial ingredients developed branch-and-cut algorithm since real-life social instances may very large. In computational study, considered variants this outperform state-of-the-art approach by orders magnitude.