Diffusion Independent Semi-Bandit Influence Maximization

We consider influence maximization (IM) in social networks, which is the problem of maximizing the number of users that become aware of a product by selecting a set of “seed” users to expose the product to. While prior work assumes a known model of information diffusion, we propose a parametrization in terms of pairwise reachability which makes our framework agnostic to the underlying diffusion model. We give a corresponding monotone, submodular surrogate function, and show that it is a good approximation to the original IM objective. We also consider the case of a new marketer looking to exploit an existing social network, while simultaneously learning the factors governing information propagation. For this, we propose a pairwise-influence semi-bandit feedback model and develop a LinUCB-based bandit algorithm. Our model-independent regret analysis shows that our bound on the cumulative regret has a better (as compared to previous work) dependence on the size of the network. By using the graph Laplacian eigenbasis to construct features, we describe a practical LinUCB implementation. Experimental evaluation suggests that our framework is robust to the underlying diffusion model and can efficiently learn a near-optimal solution.