Fair Influence Maximization in Hypergraphs

Influence maximization (IM) traditionally prioritizes maximizing influence spread, often resulting in challenges such as unfair information dissemination and social bias. To mitigate these issues, the fair influence maximization (FIM) has been introduced, focusing on minimizing the influence gap between different groups while still maximizing influence spread. However, existing FIM approaches predominantly target simple networks, neglecting the higher-order relationships that are prevalent in real-world systems. In this paper, we extend the FIM problem to hypergraphs and propose the Local Attribute and Higher-order Degree-based (LAHD) method. LAHD addresses FIM by evaluating node fairness through the local attribute distribution of neighboring nodes and node influence through higher-order degree metrics. Experimental results on both real-world and synthetic hypergraphs demonstrate that LAHD outperforms baseline methods in achieving both fairness and influence spread.