Detecting Overlapping Communities in Modularity Optimization by Reweighting Vertices

Area of Science:

• Network Science
• Computer Science
• Data Mining

Background:

• Traditional community detection algorithms struggle with overlapping community structures.
• Modularity maximization and fuzzy modularity functions are insufficient for identifying complex, overlapping communities.

Purpose of the Study:

• To address the challenge of detecting overlapping communities in real-world networks.
• To propose a novel algorithm that accurately identifies nodes belonging to multiple communities.

Main Methods:

• Formulated the overlapping community detection problem as a node weight allocation problem.
• Developed an extended modularity measure based on reweighting nodes.
• Utilized a genetic algorithm to solve the node weight allocation and detect overlapping communities.
• Implemented three refinement strategies to enhance solution quality.

Main Results:

• The proposed method successfully detected nontrivial overlapping nodes in both synthetic and real-world networks.
• The algorithm identified valuable overlapping nodes that were overlooked by existing methods.
• Experimental results demonstrate the effectiveness of the reweighting approach for overlapping community detection.

Conclusions:

• The proposed node weight allocation and genetic algorithm approach offers a significant improvement for detecting overlapping communities.
• This method enhances the accuracy and comprehensiveness of community detection in complex networks.
• The findings suggest a more robust way to analyze network structures with overlapping community memberships.