Leveraging a Hierarchical Selection of Communities in Social Networks to Maximize Diffusion

Abstract

For businesses, discovering key nodes in social networks has a high computational overhead and cost. Because of this, in the problem of influence maximization, at least key nodes with the most influence in social networks are chosen at the lowest possible cost. Numerous algorithms have been proposed to find key nodes in social networks using the technique of influence maximization under the characteristics of submodularity and monotonicity; however, these algorithms struggle with the issue of optimal influence spread because some graph characteristics are ignored when considering diffusion and influence spread in the social network. The HSMD method, which gives special weight to the structure of communities and selects key nodes hierarchically utilizing the structure of communities, has been presented as a solution to this challenge. Based on the characteristics and strategic placement of the nodes within each community, weight is given to each community in this algorithm. Seed nodes are then chosen hierarchically using BFS traversal. This method enhanced the influence spread rate and execution time of recently introduced algorithms such as CTIM, IMBC, RNR, HEDVGreedy, and k-shell.