Socially Fair Mitigation of Misinformation on Social Networks via Constraint Stochastic Optimization

Recent social networks' misinformation mitigation approaches tend to investigate how reduce by considering a whole-network statistical scale. However, unbalanced exposures among individuals urge study fair allocation of resources. Moreover, the network has random dynamics which change over time. Therefore, we introduce stochastic and non-stationary knapsack problem, apply its resolution mitigate in campaigns. We further propose generic algorithm that is robust different statistics, allowing promising impact real-world scenarios. A novel loss function ensures users. achieve fairness intelligently allocating incentivization budget knapsack, optimizing function. To this end, team Learning Automata (LA) drives allocation. Each LA associated with user learns minimize exposure performing walk state space. Our results show our LA-based method outperforms similar methods fairly influencing