MIM2: Multiple influence maximization across multiple social networks

Abstract

Influence maximization (IM) is the problem of selecting a small subset of users with the aim of maximizing influence spread to help marketers in promoting their products. None of the existing literature considers the scenario that a marketing company wants to promote multiple products in multiple networks or a network with the different channel of interactions simultaneously. Considering this scenario, we introduce multiple influence maximization across multiple social networks (MIM2) problem. This problem considers the assumption that an influential user can accept multiple products for free and non-influential users have enough purchasing power to adopt multiple promotions from their social interactions. It is also important to consider the role of overlapping users to spread the influence across networks. To address these issues, we propose a unified framework to analyze and represent the MIM2 problem. More specifically, first, we perform a mapping to couple a set of networks into a multiplex network via direct linkage strategy. Second, we propose a heuristic method to find the most influential user over multiple product diffusion multiplex networks. Third, we prove that MIM2 problem is NP-hard and expected influence spread function is sub-modular under traditional diffusion models. Finally, the experimental results show that the advantage of proposed IM problem over existing IM problems. © 2019 Elsevier B.V.