Detecting event-related changes in organizational networks using optimized neural network models.

Area of Science:

• Social Network Analysis
• Organizational Behavior Dynamics
• Computational Social Science

Background:

• Organizational external behaviors, or behavioral events, stem from internal structure and interactions.
• Existing methods for detecting changes in organizational networks often overlook the link between internal structure and external events.
• Monitoring organizational network dynamics through event-related change detection offers efficient insights into behavioral shifts.

Purpose of the Study:

• To define and detect event-related changes within organizational networks.
• To establish a method that considers the correlation between internal network structure and external organizational events.
• To develop a system for early warning and rapid response to organizational activities.

Main Methods:

• Utilized social network modeling and supervised classification for event recognition.
• Employed artificial neural network models, specifically Back Propagation Neural Networks (BPNNs).
• Optimized BPNNs using Genetic Algorithms (GAs) and Particle Swarm Optimization (PSO) for enhanced accuracy.

Main Results:

• Successfully defined and quantitatively determined event-related changes in organizational networks.
• Demonstrated the proposed method's feasibility through comparative analysis in two case studies.
• Achieved higher precision and robustness compared to existing techniques in identifying organizational events.

Conclusions:

• The proposed method effectively identifies organizational events by correlating network structures with behavioral events.
• This approach provides a robust framework for understanding and responding to organizational dynamics.
• The optimized neural network models offer a significant advancement in detecting and analyzing organizational behavioral changes.