A New Perspective to Node Influence Evaluation in Complex Network Using Subgraph Tr-Centrality

TL;DR

This paper introduces Tr-centrality, a new node influence measure based on subgraph triangle structures, validated on real-world networks, showing improved performance over existing measures.

Contribution

It proposes a novel centrality measure called Tr-centrality that leverages triangle structures and neighborhood information for node influence evaluation.

Findings

Tr-centrality outperforms existing measures on real-world networks.

The measure effectively captures local trust and influence in complex networks.

Validation across diverse network types confirms its robustness.

Abstract

There is great significance in evaluating a node's Influence ranking in complex networks. Over the years, many researchers have presented different measures for quantifying node interconnectedness within networks. Therefore, this paper introduces a centrality measure called Tr-centrality which focuses on using the node triangle structure and the node neighborhood information to define the strength of a node, which is defined as the summation of Gruebler's Equation of the node's one-hop triangle neighborhood to the number of all the edges in the subgraph. Furthermore, we socially consider it as the local trust of a node. To verify the validity of Tr-centrality [1], we apply it to four real-world networks with different densities and shapes, and Tr-centrality has proven to yield better results.