A Composite Centrality Measure for Improved Identification of Influential Users

TL;DR

This paper introduces a new composite centrality measure combining degree and k-shell indices to better identify influential users in social networks, outperforming existing methods in accuracy and monotonicity.

Contribution

It proposes a novel composite centrality measure that integrates degree and k-shell indices, improving influence ranking accuracy over existing measures.

Findings

The proposed measure outperforms state-of-the-art methods in experiments.

It achieves higher accuracy and better monotonicity in influence ranking.

Experimental results validate the effectiveness of the composite measure.

Abstract

In recent years, the problem of identifying the spreading ability and ranking social network users according to their influence has attracted a lot of attention; different approaches have been proposed for this purpose. Most of these approaches rely on the topological location of nodes and their neighbours in the graph to provide a measure that estimates the spreading ability of users. One of the most well-known measures is k-shell; additional measures have been proposed based on it. However, as the same k-shell index may be assigned to nodes with different degrees, this measure suffers from low accuracy. This paper is trying to improve this by proposing a composite centrality measure in that it combines both the degree and k-shell index of nodes. Experimental results and evaluations of the proposed measure on various real and artificial networks show that the proposed measure outperforms other state-of-the-art measures regarding monotonicity and accuracy.