A hybrid approach in opinion leaders selection using African vultures optimization and hunger games search algorithms

Depending on the actions and reactions in social networks, the opinions of users may have an impact on each other. Nodes who can impact others’ opinions are defined as opinion leaders. In this paper, we formulate Opinion Leader Selection (OLS) as an optimization problem, for this purpose, African vultures optimization algorithm (AVOA) and Hunger games search (HGS) algorithms are utilized to choose opinion leaders and optimize the results. The proposed method uses the topological characteristics of the network and the combination of relationships between users, also a model that includes four activities, stability, novelty, and popularity features to calculate the fitness function and other required parameters. The Social Network Analysis (SNA) parameters, SIR, and OSim models are used to investigate the proposed method. Based on the findings of the evaluation, the proposed method provides for the Bitcoin alpha dataset (0.88, 0.7, 0.97, 0.81) percent and the Bitcoin OTC dataset (0.76%, 0.64, 0.96, 0.77) percent for accuracy, recall, precision, and F1-score respectively. Also, on the SIR and OSim models analysis, compared to the other 8 approaches, the proposed method gives a better-spreading ability interval and results.