Computing Influential Nodes Using the Nearest Neighborhood Trust Value and PageRank in Complex Networks.

Area of Science:

• Network Science
• Data Mining
• Computational Social Science

Background:

• Identifying influential nodes is crucial for understanding information diffusion in complex networks.
• Existing centrality measures (e.g., PageRank, betweenness) often have high time complexity, limiting their use in large-scale networks.
• There is a need for centrality measures that leverage both network structure and node attributes.

Purpose of the Study:

• To propose a novel centrality measure, Nearest Neighborhood Trust PageRank (NTPR), for identifying influential nodes.
• To develop a method that considers structural attributes of neighbors and trust values.
• To evaluate the effectiveness of NTPR in real-world networks.

Main Methods:

• Developed the Nearest Neighborhood Trust PageRank (NTPR) measure.
• NTPR incorporates degree ratio, node similarity, and trust values of neighbors and nearest neighbors.
• Evaluated NTPR on various real-world networks using SIR and independent cascade models.

Main Results:

• The proposed NTPR method effectively identifies influential nodes in complex networks.
• NTPR demonstrated superior performance in maximizing information influence compared to existing basic centrality measures.
• The method's reliance on local and attribute-based information addresses scalability issues.

Conclusions:

• NTPR offers an efficient and effective approach for computing influential nodes in large-scale networks.
• The method's ability to integrate structural and trust-based attributes enhances its applicability in diverse network analysis tasks.
• This research contributes a valuable tool for applications like viral marketing and rumor control.