Micro-blog user community discovery using generalized SimRank edge weighting method.

Area of Science:

• Network analysis
• Social network analysis
• Data mining

Background:

• Community discovery is crucial for understanding network structures.
• Edge weighting can enhance community discovery accuracy.
• Existing methods may not fully capture complex network relationships.

Purpose of the Study:

• To propose a novel edge weighting method balancing local and global information for improved community discovery.
• To introduce a new evaluation metric, communication rate, for assessing community division quality.
• To adapt existing algorithms for weighted networks and evaluate performance on real-world data.

Main Methods:

• A novel edge weighting method considering shared neighbors and interpersonal significance.
• Adaptation of the fast Newman algorithm for weighted networks.
• Evaluation using a communication rate metric and comparison with modularity Q.

Main Results:

• The proposed method effectively balances local and global weighting, adapting to diverse real-world networks.
• The communication rate metric provides a better measure of user interaction relations than modularity Q.
• Experimental results on Sina Weibo data show the method outperforms conventional weighting techniques.

Conclusions:

• The novel edge weighting method offers a robust and effective approach to network community discovery.
• The proposed evaluation metric enhances the understanding of community structures in social networks.
• This work contributes to advancing social network analysis through improved weighting and evaluation strategies.