Missing Link Prediction using Common Neighbor and Centrality based Parameterized Algorithm.

Area of Science:

• Network Science
• Data Mining
• Computational Social Science

Background:

• Real-world complex networks are dynamic and often fragmented.
• Link prediction is crucial for understanding network evolution and for applications like friend recommendations and item suggestions.
• Existing methods face challenges in accurately predicting new links in complex network structures.

Purpose of the Study:

• To propose a novel link prediction algorithm named Common Neighbor and Centrality based Parameterized Algorithm (CCPA).
• To enhance the accuracy of predicting new links in complex networks.
• To provide a robust method for applications in social networks and recommender systems.

Main Methods:

• Developed the Common Neighbor and Centrality based Parameterized Algorithm (CCPA).
• Utilized the Area Under the receiver operating characteristic Curve (AUC) as the primary evaluation metric.
• Conducted extensive experiments on eight diverse real-world network datasets.

Main Results:

• The CCPA algorithm demonstrated superior performance compared to eight benchmark algorithms.
• Evaluations across multiple datasets confirmed the effectiveness of the proposed approach.
• The CCPA algorithm showed significant improvements in link prediction accuracy.

Conclusions:

• The proposed CCPA algorithm offers an effective solution for link prediction in complex networks.
• CCPA provides a valuable tool for enhancing recommendations in social and e-commerce platforms.
• The algorithm's performance validates its potential for real-world network analysis.