On Bharathi–Kempe–Salek conjecture for influence maximization on arborescence

Abstract

Bharathi et al. (WINE, pp 306---311, 2007) conjectured that the influence maximization problem is NP-hard for arborescence directed into a root. In this note, we show that this conjecture is not true for deterministic diffusion model and linear threshold (LT) model, that is, there exist polynomial-time algorithms for the influence maximization problem in those two models on arborescence directed into a root. This means that if the conjecture in the independent cascade (IC) model is true, then it would give an interesting difference between the IC model and the LT model.