|
Abstract
|
Community detection in dynamic social networks is crucial for understanding evolving relationships and structures. While traditional methods struggle with modern networks' complexity, deep learning techniques like Graph Attention Networks (GATs) and multi-objective optimization (MOO) offer promising solutions by capturing intricate node relationships and balancing conflicting goals. Current research explores combining these approaches to enhance the precision of community detection, addressing limitations in accuracy, efficiency, and adaptability. This research aims to analyze existing algorithms, propose a novel method integrating GAT and MOO, and investigate the challenges in dynamic networks.
|