I am wondering is there an implementation for node prediction of GraphSage for heterogenous graphs (one type of node and two types of edges)? Similar to this post, but seems not confirmed.
I don’t think currently DGL has an example for node classification on heterogeneous graphs with GraphSAGE.