Fairness-Aware Influence Blocking Maximization in Social Networks

Nowadays, social media continually influences people’s opinions and behaviors. The spread of unchecked rumors has a detrimental impact on society. It is crucial to develop effective strategies to block the dissemination of rumors as much as possible. However, current approaches for influence blocking overlook the fairness among diverse communities, which are fundamental structures in social networks, leading to a disproportionate exclusion of marginalized communities from intervention benefits and resulting in fairness disparities. This article addresses such a critical issue in influence blocking and proposes a fairness-aware blocking strategy that balances the effectiveness of influence blocking and fairness. A sampling-based approximate algorithm called fairness-aware influence blocking maximization based on martingales (FIBMM), which is applicable to large networks, is presented. We conduct a series of experiments on six real networks, showing that the FIBMM can effectively reduce fairness disparities between diverse communities while maintaining blocking effectiveness.