An effective Fuzzy Sign-aware Influence Maximization method in complex networks using an adaptive Improved Grey Wolf Optimizer with dynamic user interactions

The widespread use of social networks worldwide has significantly impacted personal lives and various industries. One prominent area affected is marketing and advertising, where viral marketing has become an effective strategy for rapidly reaching large audiences. However, this advancement has also brought notable challenges, such as the influence maximization (IM) problem. IM involves identifying a small set of individuals who can maximize influence propagation across a social network. Existing methods for solving this NP-hard optimization problem still fail to meet the performance required for real-world applications. This shortcoming arises primarily from the inability of these methods to effectively focus the search towards optimal regions to identify final solutions and their failure to comprehensively capture the dynamic interactions and social influences in complex user relationships with varying levels of trust. This paper proposes a new method called Fuzzy Sign-aware Influence Maximization using an adaptive Improved Grey Wolf Optimizer (FSIMI-GWO) to tackle these limitations in complex signed networks. The FSIMI-GWO introduces a fitness function that incorporates fuzzy logic to effectively apply diverse user relationships through appropriate weight assignments. Furthermore, this function adapts the method's search strategy to find promising areas according to dynamic changes in user interactions. The proposed method is evaluated against several well-known and recently introduced methods for influence maximization on real-world networks. The evaluation is conducted using developed signed variants of the popular cascade diffusion model. Extensive experiments confirm that the proposed method attains superior influence propagation compared with the baselines.