Discrete particle swarm optimization for identifying community structures in signed social networks

Area of Science:

• Network Science
• Computational Social Science
• Artificial Intelligence

Background:

• Community structure is a key feature of complex networks.
• Identifying communities is often framed as an optimization problem.
• Existing optimization methods like Particle Swarm Optimization (PSO) are primarily designed for continuous, not discrete, problems.

Purpose of the Study:

• To develop a novel discrete Particle Swarm Optimization (PSO) algorithm.
• To adapt PSO for the specific challenge of identifying community structures in signed networks.
• To evaluate the proposed algorithm's effectiveness against existing state-of-the-art methods.

Main Methods:

• A novel discrete PSO algorithm was designed.
• Particle states were redefined for discrete optimization scenarios.
• Particle updating rules were reformulated based on signed network topology.
• The algorithm was tested on both synthetic and real-world signed networks.

Main Results:

• The proposed discrete PSO algorithm demonstrated effectiveness in identifying community structures.
• Experimental results showed competitive or superior performance compared to three state-of-the-art approaches.
• The method proved promising for analyzing complex systems represented by signed networks.

Conclusions:

• The novel discrete PSO algorithm is a viable and effective tool for community detection in signed networks.
• Adapting PSO for discrete optimization enhances its applicability to network science problems.
• The proposed method offers a promising advancement in the analysis of complex network structures.