Optimal seed node selection method for LTIS model

Abstract

The problem of seed node selection to maximize the influence diffusion is one of the most pivotal problems related to applications like marketing campaigns, spreading awareness, and so forth. Most of the existing work focuses on single-step influence diffusion where the seed nodes are selected in one go to trigger the diffusion process, and the influence spread is mostly studied under progressive models. This article proposes a new model named the LTIS model, where the diffusion process can take place in multiple steps. In this model, a node that gets influenced in one step can again go to an uninfluenced state in subsequent steps. This article also proposes an optimal seed node selection method named advantage-of-seed-node method to find the seed node-set. This proposed method accurately computes the advantage of the seed node to find the influence spread of each node. A comparative analysis of the proposed method in terms of activated nodes from a selected seed node-set under the proposed LTIS model outperforms many existing seed set selection schemes.