Network alignment in multilayer dynamic social network using the information diffusion dynamics

Abstract Multilayer dynamic social network alignment aims to align different social networks based on their common users, a topic that has gained significant attention in academia and industry. However, most existing studies treat social networks as static, overlooking the dynamic characteristics that play a crucial role in individual behavior patterns. Dynamic multilayer social networks consist of multiple distinct layers that evolve over time, and their structure presents substantial challenges for understanding internal relationships. Without proper network alignment, issues such as ambiguous node correspondence and unclear interaction patterns arise, which negatively impacts the accuracy of subsequent data analysis and mining tasks. To address these issues, we propose a novel multilayer dynamic social network alignment framework that captures node correspondences across layers through information diffusion. Experimental results demonstrate that our framework performs exceptionally well on both simulated networks and real-world datasets, demonstrating its potential applications in social influence analysis, community detection, and recommendation systems.