Epidemic Model-based Network Influential Node Ranking Methods: A Ranking Rationality Perspective

TL;DR

This paper surveys influential node ranking methods based on epidemic models, analyzing their capability and correctness in ranking nodes across various network types, and offers insights for better method selection and future research directions.

Contribution

It systematically categorizes INRMs, introduces evaluation metrics for ranking rationality, and analyzes their effectiveness across different network types, filling a gap in existing research.

Findings

Classified 7 categories of INRMs and 4 network types

Defined Capability and Correctness metrics for evaluation

Provided insights into the ranking rationality of INRMs

Abstract

Most recent surveys and reviews on Influential Node Ranking Methods (INRMs) hightlight discussions on the methods' technical details, but there still lacks in-depth research on the fundamental issue of how to verify the considerable influence of these nodes in a network. Compared to conventional verification models such as cascade failure and linear threshold, the epidemic model is more widely used. Accordingly, we conducted a survey of INRM based on epidemic model on 81 primary studies and analyzed their Capability and Correctness which we defined in our work. Our study categorized 4 types of networks used by INRM, classified 7 categories of INRMs for analyzing the networks and defined 2 evaluation metrics set of Capability and Correctness for evaluating INRM from Ranking Rationality Perspective. We also discussed particular real-world networks that were used to evaluate INRM and the Capability and Correctness of different INRMs on ranking nodes in specific networks. This is, as far as we know, the first survey aimed at systematically summarizing the Capability and Correctness of INRM. Our findings can assist practitioners and researchers in choosing and comparing INRMs and identifying research gaps.