Review for NeurIPS paper: Co-exposure Maximization in Online Social Networks

Summary and Contributions: The paper considers an extension to the influence maximization problem where two campaigns co-exist in the network and the objective is to allocate a set of seed-nodes to each campaign, under cardinality constraints, such that the expected number of nodes that are exposed to both campaigns is maximized. The diffusion model is assumed to be the Independent Cascade model. The problem is well-motivated, as it is clear that the problem of polarization exists in social networks and media, and the network owners have incentives to take action in reducing the polarization among their users. The theoretical findings in the paper are very interesting and novel, and the ideas explored can be applicable to other problems as well. I enjoyed reading the paper.