This tutorial gives an overview of some of the basic work that has been done over the last five years on the application of deep learning techniques to data represented as graphs. Convolutional neural networks and transformers have been instrumental in the progress on computer vision and natural language understanding. Here we look at the generalizations of these methods to solving problems where the data is represented as a graph. We illustrate this with examples including predicting research topics by using the Microsoft co-author graph or the more heterogeneous ACM author-paper-venue citation graph. This later case is of interest because it allows us to discuss how these techniques can be applied to the massive heterogeneous Knowledge networks being developed and used by the search engines and smart, interactive digital assistants. Finally, we look at how knowledge is represented by families of graphs. The example we use here is from the Tox21 dataset of chemical compounds and their interaction with important biological pathways and targets.

The full tutorial is on our other site: https://cloud4scieng.org/2020/08/28/deep-learning-on-graphs-a-tutorial/