It's a generalization of a fully connected layer, where you give the network an adjacency matrix for data routing between nodes. This allows for designing sparse data flows and learning relations between pairs of input elements.

Graph NN layer: f(X, A) = g(AXW), where g is an activation

Fully connected layer: f(X) = g(XW), like GNN where the adj. matrix A=I

To put it in context a few years ago GNNs became a hot field. They are very similar in a way to transformers because both do pairwise interactions between elements, the difference being that GNNs use explicit and transformers implicit graphs.

Is this a simplified interpretation of GNNs? because I couldn't find a reference for that.

EDIT: found it! https://arxiv.org/pdf/1609.02907.pdf

A routine automation task is information extraction from any kind of document with a visual layout - receipts, PDFs, web pages. [1] [2]

[1] : https://nanonets.com/blog/information-extraction-graph-convo...

[2] : https://arxiv.org/pdf/1903.11279.pdf

Traffic prediction https://deepmind.com/blog/article/traffic-prediction-with-ad...

The transformer is a fully connected graph neural network. Read https://arxiv.org/abs/2004.08483 for further ideas.

Chemical structures can be described as graphs.

They’re seeing some interest in neuroscience as the brain can be represented with a graph structure.

fraud detection

expanding on that: people used to do stuff like pagerank etc. ahead of time and decorate their tabular data with the results: entities, events, whatever for social, fraud, security, customer journey, whatever. that meant a phase separation between graph analytic enrichments and learning. bringing the non-local graph reasoning to the learning phase in a way that isn't slow enables closing the loop.

The early graphsage stuff was, afaict, proven for generic social recommendors, but most gnn's I see seem pretty custom (e.g., deepmind's protein folding solution), esp. when not prohibitively slow. It sounds like more generic use is becoming practical w/ these libs, and esp. interesting to me, the latest NIPS had graph transformers papers, which brings another level of practically here. Not sure if DGL & friends have those yet..