Provably Efficient Reinforcement Learning for Online Adaptive Influence Maximization
This work addresses the challenge of adaptively selecting seed nodes for influence spread in social networks, which is incremental as it extends prior non-adaptive methods to a sequential, feedback-driven setting.
The paper tackles the problem of online adaptive influence maximization in social networks with unknown models by formulating it as an MDP and using a model-based reinforcement learning approach, achieving a regret bound of O(√T) and demonstrating efficiency in synthetic network evaluations.
Online influence maximization aims to maximize the influence spread of a content in a social network with unknown network model by selecting a few seed nodes. Recent studies followed a non-adaptive setting, where the seed nodes are selected before the start of the diffusion process and network parameters are updated when the diffusion stops. We consider an adaptive version of content-dependent online influence maximization problem where the seed nodes are sequentially activated based on real-time feedback. In this paper, we formulate the problem as an infinite-horizon discounted MDP under a linear diffusion process and present a model-based reinforcement learning solution. Our algorithm maintains a network model estimate and selects seed users adaptively, exploring the social network while improving the optimal policy optimistically. We establish $\widetilde O(\sqrt{T})$ regret bound for our algorithm. Empirical evaluations on synthetic network demonstrate the efficiency of our algorithm.