GAC: A Deep Reinforcement Learning Model Toward User Incentivization in Unknown Social Networks

In recent years, providing incentives to human users for attracting their attention and engagement has been widely adopted in many applications. To effectively incentivize users, most incentive mechanisms determine incentive values based on users' individual attributes, such as preferences. These approaches could be ineffective when such information is unavailable. Meanwhile, due to the budget limitation, the number of users who can be incentivized is also restricted. In this light, we intend to utilize social influence among users to maximize the incentivization. By directly incentivizing influential users in the social network, their followers and friends could be indirectly incentivized with fewer incentives or no incentive. However, it is difficult to identify influential users beforehand in the social network, as the influence strength between each pair of users is typically unknown. In this work, we propose an end-to-end reinforcement learning-based framework, named Geometric Actor-Critic (GAC), to discover effective incentive allocation policies under limited budgets. More specifically, the proposed approach can extract information from a high-level network representation for learning effective incentive allocation policies. The proposed GAC only requires the topology of the social network and does not rely on any prior information about users' attributes. We use three real-world social network datasets to evaluate the performance of the proposed GAC. The experimental results demonstrate the effectiveness of the proposed approach.