Heuristics for the Influence Maximization Problem on Hypergraphs

While the Influence Maximization (IM) problem has been widely analyzed on graph topologies, with plenty of algorithms and heuristics proposed, very limited interest has been devoted to the study of this problem on more complex, but much more expressive, hypergraph topologies. In this extended abstract, we address this issue by illustrating three families of heuristics for this problem: the first is based on node importance measures appropriately defined on hypergraphs, based both on its topological features, and on cooperative game theory concepts; while the remaining ones cast the problem within the well-known metaheuristic frameworks of Hill Climbing and Evolution Strategy, respectively. We experimentally evaluated these approaches, and we show that they often outperform the best-proposed algorithms in the state of the art.