Fairness-aware influence maximization: A novel Learning Automata-based approach

Abstract

In the current era, due to the widespread availability of the Internet and the spread of social media, there has been a lot of increase in social influence through the Internet. There are a lot of influencers who help spread the popularity of a product or opinion through their word of mouth. In such a scenario, targeting the right influencers to spread the information becomes an important problem. Influence maximization (IM) tackles the problem of finding a seed set of influencers in a graph which leads to the maximum spread of information. The state-of-the-art algorithms that aim to solve IM problems are unable to choose a seed set that is fairly chosen. Existing algorithms that focus on fairness do not give large activation. Hence, to solve the IM problem with regard to the fairness of the seed set, this work proposes a Learning Automata-based Sine Cosine Algorithm with Levy Flights to heuristically search for a seed set. In the proposed work, fairness indicates the fair distribution of influence among different communities/groups. The proposed SCA algorithm uses Levy Flights for the exploration and integrates the Learning Automata (LA) to make it adaptive. This work uses Pareto optimality that ranks the solutions to optimize the influence and fairness. The experiments have been conducted on eight datasets, and fairness is evaluated using four metrics. The results analysis shows the highest activated nodes and competitive and even higher fairness within adequate time as compared to benchmark algorithms. © 2025 Elsevier Ltd