DQ4FairIM: Fairness-Aware Influence Maximization Using Deep Reinforcement Learning

Abstract

The influence maximization (IM) problem aims to select a set of seed nodes within a given budget to maximize the spread of influence in a social network. However, real-world social networks have several structural inequalities, such as dominant majority groups and underrepresented minority groups. If these inequalities are not considered while designing IM algorithms, the outcomes might be biased, disproportionately benefiting majority groups while marginalizing minorities. In this work, we address this gap by designing a fairness-aware IM method using deep reinforcement learning (RL) that ensures equitable influence outreach across all communities, regardless of protected attributes. Fairness is incorporated using the maximin fairness objective, which prioritizes improving the outreach of the least-influenced group, pushing the solution toward an equitable influence distribution. We propose a novel fairness-aware deep RL method, called deep $\boldsymbol{Q}$-learning for fair influence maximization (DQ4FairIM), that learns a seeding policy to maximize the expected number of influenced nodes, ensuring that minority groups are not disproportionately excluded. The IM problem is formulated as a Markov decision process and solved using deep Q-learning combined with Structure2vec network embeddings. The main benefit of using RL for this problem is its generalization ability, in which a policy learnt on a specific problem instance can be generalized to new problem instances without retraining the model from scratch. Extensive experiments on synthetic benchmarks and real-world networks demonstrate that DQ4FairIM consistently achieves higher fairness and a better fairness–performance trade-off compared to both fairness-agnostic and existing fairness-aware baselines, while retaining strong generalization across varying network sizes and seed budgets.