Threshold Benefit for Groups Influence in Online Social Networks

Abstract

The influence maximization (IM) is an optimization problem in the information propagation and social network analysis, which has the goal of finding a seed set that can influence largest number of users. There have been many studies on the IM problem, but most of them focus on maximizing influence effects based on individuals rather than on groups of users. In this paper, we studied a novel problem, the Threshold Benefit for Groups Influence (named \({\mathsf {TGI}}\)), defined as follows: given a social network G, a set of K target groups \({\mathcal C} =\{C_1, C_2, \ldots , C_K\}\) and a threshold \(T>0\), \({\mathsf {TGI}}\) asks us to find a seed set S in G with the minimum cost so that the benefit gained from the influence of the groups in C is at least T. We experimentally implement our proposed algorithm on real social network datasets diversity. It shows efficiency compared to other state-of-the-art IM algorithms in terms of cost and running time.