Attributed graphs are a powerful tool to model real life systems which exist in many domains such as social science, biology, etc. Online social networks have a significant effect on human society and have become an important research topic for maintaining the integrity of the common social understanding. This includes combating disinformation and sensing predispositions that can disrupt peaceful discourse. Studying networks often involves labeling nodes into representative groups. Simple machine learning approaches or community detection have limitation in their capability to make use of both network topology and node features. Graph Neural Networks (GNNs) provide an efficient framework combining both sources of information to produce accurate node classification. In this work, we study the application of two variants of GNNs, namely Simple Graph Convolution (SGC) and its extension on a social network dataset as well as a comprehensive set of synthetic attributed graphs with varying network topology. The SGC provides fast and efficient framework while its extension improves interpretability of the result by highlighting key node features determining the class characteristics.

PDF