Overlapping community-based fair influence maximization under a multi-transformation optimization algorithm

Fair Influence Maximization (FIM) aims to ensure an equitable spread of influence across different groups in a social network while maximizing overall reach. By incorporating fairness constraints and leveraging community structures, FIM mitigates bias and enhances the effectiveness of influence propagation. Despite its importance, optimizing FIM objectives is computationally challenging, and existing methods often neglect the overlapping nature of community structures. Also, selecting the most suitable transformation in advance is challenging, as different transformations lead to varying search behaviors for evaluating influence spread. To address these issues, this study introduces Overlapping Community-based FIM under a Multi-Transformation Optimization algorithm (OCMTO). OCMTO considers both overlapping and non-overlapping nodes within communities to balance influence spread and fairness. It employs a multi-transformation evolutionary framework that optimizes multiple transformations simultaneously, ensuring scalability and convergence. The algorithm adapts to different transformations without the need for manual selection, allowing it to efficiently explore the solution space. Meanwhile, OCMTO estimates relationships between transformations based on individual overlap. Experimental results on real-world datasets show that OCMTO outperforms state-of-the-art methods in scalability and fairness, demonstrating its practical potential for applications such as digital marketing campaigns, public health interventions, and political influence dissemination. Specifically, the proposed algorithm demonstrates a 9.2% improvement in average influence spread over the best existing method, while effectively addressing the trade-offs between influence, fairness, and complexity.