Key Node Ranking in Complex Networks: A Novel Entropy and Mutual Information-Based Approach

Area of Science:

• Complex network analysis
• Network science
• Information theory

Background:

• Identifying key nodes is crucial for understanding complex network behavior.
• Existing methods often rely on limited local information and struggle with global network properties.
• Many current approaches are computationally expensive (NP-hard).

Purpose of the Study:

• To propose a novel and efficient method for identifying key nodes in complex networks.
• To develop a centrality approach that incorporates both topological and digital network characteristics.
• To address the limitations of existing methods by leveraging a wider range of network information.

Main Methods:

• Developed an entropy and mutual information-based centrality (EMI) approach.
• Integrated local and global network information for node importance assessment.
• Utilized mutual information to refine centrality calculations and overcome existing flaws.

Main Results:

• The proposed EMI approach demonstrated superior performance in identifying key nodes.
• Experiments on real-world networks confirmed the effectiveness and efficiency of EMI.
• EMI successfully captured a broader spectrum of information compared to traditional methods.

Conclusions:

• The EMI-based centrality approach offers a significant advancement in key node identification.
• This method provides a more comprehensive and accurate assessment of node importance in complex networks.
• EMI presents a computationally efficient and effective solution for network analysis challenges.