A balanced method for budgeted influence maximization

Abstract

With the flourish ofWeb-based large Online Social Networks (OSNs), people on OSNs can easily yield influence on others. Finding how the influence spreads and maximizing influence spread within OSNs have been extensively studied. State-of-the-art researches suffer two defects: (a) need to acquire the topological structure of the network, which is impractical for the continuously changing networks in real life and thus can not balance very well between influence spread and running time; (b) assign the same cost for every node in OSN which cannot reflect the reality. To solve these problems we firstly propose PageRank Based Cost (PRBC) model to assess the cost of nodes in OSN according to their importance (influence); secondly we present Budgeted Random Maximal Degree Neighbor (BRMDN) algorithm by exploiting the scale free property. Results from extensive experiments show that BRMDN can well balance influence spread and running time.