Competitive influence maximization in biased community networks: An adaptive algorithmic approach

Influence maximization is a widely recognized problem that aims to identify a limited set of influencers to maximize the coverage of influence propagation in a social network. Several studies have analyzed the problem in networks with diverse community structures. However, these studies generally consider communities to be "neutral" or "unbiased," whereas many real-world communities are "biased"–significantly influenced or even directly shaped by pioneering products. This biases introduce a new competitive influence maximization problem in networks with biased community structures from the followers’ perspective. Therefore, we propose a new propagation model that accounts for the "bias" structure present in real-world communities. Traditional influence maximization algorithms often suffer from slow convergence, limited spread, and local optima in biased communities. To address these challenges, we develop an enhanced genetic algorithm integrated with simulated annealing, which improves convergence properties and parametric adaptability. Simulated experiments demonstrate that, compared to existing algorithms, our proposed method achieves better influence maximization effects and shows strong potential for real-world applications.