Topology-Aware Graph Neural Network with Spatiotemporal Encoding and Contrastive Learning for Scalable Dynamic Network Representation
The implications of the proposed TAGNN framework on such applications as anomaly detection, social media analytics, and traffic forecasting are that the proposed framework provides an end-to-end solution that promotes accuracy, efficiency, and temporal consistency in dynamic graph learning.
Wassan Hayale, R. S. Ali, Raghda Abd Ul Rab Abd Ul Hasan et al.
· Mesopotamian Journal of Big... · 0 citations