A Primer on Temporal Graph Learning

For example, the interaction of different users on a social media platform can be represented as graphs. Similarly, citation networks capture the scholarly interdependence of academic papers through citations, offering insights into the knowledge landscape. Biological networks, encompassing interactions among bio-molecules, genes, and proteins, leverage graphs to unravel the complexities of biological systems. Building upon the accomplishments of neural networks in processing Euclidean data, graph neural networks(GNNs) are specifically crafted to operate seamlessly with graph-structured data. The foundations of GNNs can be traced to graph signal processing (GSP), which was developed to impart meaning to signal processing operations on graphical data.