A biased random key genetic algorithm applied to target set selection in viral marketing

Abstract

The Target Set Selection (TSS) problem is an NP-hard combinatorial optimization problem with origins in the field of social networks. There are various problem variants, all dealing with finding a smallest subset of vertices of a graph such that their influence is propagated to all nodes of the graph under a specific diffusion model. Despite the practical relevance of the problem, most existing research efforts have focused on theoretical properties restricted to certain classes of graphs. The richness in terms of theoretical results is in contrast to the scarceness of research aiming at efficiently solving the TSS problem. In this work we propose a Biased Random Key Genetic Algorithm (BRKGA) for solving the TSS problem in large-scale social networks. We consider the problem in combination with the Linear Threshold diffusion model. The obtained results show that our approach outperforms a recent heuristic from the literature.