Abstract:In recent years, epidemic modeling in complex networks has found many applications, including modeling of information or gossip spread in online social networks, modeling of malware spread in communication networks, and the most recent model of the COVID-19 pandemic. If the information disseminated is accurate, for example, maximizing its distribution is desirable, whereas if it is a rumor or a virus, its spread should be minimized. In this context, it is very important to identify super-spreaders that maximize or minimize propagation. Lately, studies for detecting super-spreaders have gained momentum. Most of the studies carried out aim to distinguish the influences of nodes under a specific propagation model (such as SIR) using network centrality measures and subsequently, to rank the nodes accordingly. However, in this study, we developed an algorithm that approximates the expected influence of nodes under the popular SIR model. By considering the behavior of the SIR model and only the shortest paths between nodes, the algorithm ranks the nodes according to this approximated value. Our developed algorithm is named the Expected Value Estimation (EVE). We compared the performance of EVE, using different SIR settings on real datasets, with that of many current well-known centrality measures. The experimental studies demonstrated that the solution quality (ranking capability) of EVE is superior to that of its competitors.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to estimate the influence of nodes more accurately in complex networks, especially in the case of using the Susceptible - Infected - Recovered (SIR) model. Most existing studies mainly rely on network centrality measures to distinguish the influence of nodes, while ignoring the dynamic characteristics of the propagation model. This leads to the fact that in some cases, these methods may not be able to accurately assess the actual influence of nodes. To solve this problem, the author developed an algorithm called Expected Value Estimation (EVE). The core idea of this algorithm is that the probability that a node influences another node through the shortest path is higher than through other paths, because the influence probability decays exponentially as the path length increases. Therefore, by only calculating the shortest path and the dynamic characteristics of the propagation model, the influence of nodes can be quickly estimated. ### Main contributions 1. **Considering SIR dynamics**: EVE considers not only the position of nodes in the network, but also the dynamic characteristics of the SIR model. 2. **Directly measuring influence**: Unlike centrality measures that indirectly distinguish the influence of nodes, EVE directly measures the approximate influence of nodes under the SIR model. 3. **Efficient implementation**: The implementation of the EVE algorithm is fast and easy, and the main computational cost lies in the calculation of the shortest path. 4. **Experimental verification**: Experiments on multiple real and synthetic datasets show that EVE outperforms existing centrality measure methods in terms of node influence ranking. ### Experimental results - **Kendall rank correlation coefficient**: EVE performs best in six experiments, and its Kendall’s tau value is close to or higher than 0.8. - **Monotonicity**: The monotonicity value of EVE is 1 in three experiments and is also very close to 1 in other experiments. - **Number of Top x% matching nodes**: In the top 3% and top 5% node rankings, EVE performs best in four experiments. ### Limitations - **Dependence on SIR model**: The EVE algorithm depends on the SIR model and is not applicable to other propagation models. - **High - density networks**: In high - density networks, because there may be multiple paths between nodes, EVE may underestimate the influence of nodes. In conclusion, this paper proposes a new method to more accurately estimate the influence of nodes in complex networks by combining the shortest path and the dynamic characteristics of the SIR model. The experimental results show that EVE performs excellently on multiple datasets and outperforms existing centrality measure methods.

Estimating the Expected Influence Capacities of Nodes in Complex Networks under the Susceptible-Infectious-Recovered (SIR) Model

A novel weight neighborhood centrality algorithm for identifying influential spreaders in complex networks

Lexical sorting centrality to distinguish spreading abilities of nodes in complex networks under the Susceptible-Infectious-Recovered (SIR) model

Comparative evaluation of community-aware centrality measures

Modeling Epidemics Spreading on Social Contact Networks

Identifying influential spreaders based on indirect spreading in neighborhood

Comparing Community-aware Centrality Measures in Online Social Networks

A New Measure of Identifying Influential Nodes: Efficiency Centrality

Identifying influential spreaders in complex networks for disease spread and control

Combined Centrality Measures for an Improved Characterization of Influence Spread in Social Networks

Towards Identifying Influential Nodes in Complex Networks Using Semi-Local Centrality Metrics

An Improved Evaluating Method of Node Spreading Influence in Complex Network Based on Information Spreading Probability

Unveiling Influence in Networks: A Novel Centrality Metric and Comparative Analysis through Graph-Based Models

Identifying Influential Spreaders in Complex Network Based on the Node's Weight and Spreading Probability

The Impact of Information Dissemination Strategies to Epidemic Spreading on Complex Networks

An Environment Aware Epidemic Spreading Model and Immune Strategy in Complex Networks.

Identifying Influential Spreaders in Complex Networks Based on Entropy Weight Method and Gravity Law

Identifying Influential Nodes in Social Networks: Centripetal Centrality and Seed Exclusion Approach

Contagion-Preserving Network Sparsifiers: Exploring Epidemic Edge Importance Utilizing Effective Resistance

Identifying the Influential Latent Edges for Promoting the Co-SIR Model.

SIVR Spreading Model with Anti-Virus Factor in Complex Networks