Identifying Influential Nodes in Complex Networks Based on Multiple Local Attributes and Information Entropy.

Area of Science:

• Network Science
• Graph Theory
• Data Mining

Background:

• Identifying influential nodes is crucial for understanding complex networks.
• Global attribute-based methods struggle with large-scale networks due to high time complexity.
• Multi-attribute approaches outperform single-attribute methods for node influence evaluation.

Purpose of the Study:

• To propose a novel method for identifying influential nodes in complex networks.
• To enhance node influence evaluation by incorporating multiple local attributes and neighborhood information.
• To address the limitations of existing methods in terms of time complexity and accuracy.

Main Methods:

• Developed a new multiple local attributes-weighted centrality (LWC) method.
• Combined degree and clustering coefficient for direct influence measures.
• Incorporated one-step and two-step neighborhood information for indirect influence measures.
• Utilized information entropy to weight four influence measures (degree, clustering coefficient, two-hop degree, two-hop clustering coefficient).

Main Results:

• The proposed LWC method effectively identifies influential nodes in real-world networks.
• Experimental comparisons show superior discrimination capability and accuracy compared to five well-known methods.
• The method demonstrates good performance across four different real-world network datasets.

Conclusions:

• The LWC method offers an efficient and accurate approach for identifying influential nodes in large-scale complex networks.
• Integrating multiple local attributes and neighborhood information significantly improves node influence identification.
• The proposed method provides a valuable tool for network analysis and understanding.