TaCD: Team-Aware Community Detection Based on Multi-View Modularity

Area of Science:

• Network science
• Social network analysis
• Data mining

Background:

• Community detection is crucial in network science.
• Existing methods often ignore vital team information in social networks.
• Team structures significantly influence community formation and evolution.

Purpose of the Study:

• To propose a novel community detection algorithm that integrates team information.
• To develop a multi-view network model incorporating user interactions and team structures.
• To introduce a new dataset for evaluating team-aware community detection.

Main Methods:

• Constructed a multi-view network with user interaction (user view) and team information (team view).
• Utilized Jaccard similarity for cross-view coupling to measure consistency.
• Employed multi-view modularity and the Generalized Louvain approach for optimization.

Main Results:

• The proposed Team-aware Community Detection (TaCD) algorithm effectively identifies communities.
• TaCD demonstrates superior performance compared to existing algorithms on benchmark datasets.
• The new SCHOLAT dataset facilitates research in team-aware community detection.

Conclusions:

• Integrating team information enhances community detection accuracy in social networks.
• TaCD provides a robust framework for uncovering complex community structures.
• The developed algorithm and dataset contribute significantly to the field of network science.