Fig. 1
From: edge2vec: Representation learning using edge semantics for biomedical knowledge discovery
An illustrative pipeline of edge2vec. a a heterogeneous network with three types of nodes and two types of edges, colored by types. b EM framework to optimize an edge-type transition matrix M and generate node random walks as well as related edge-type corpus. c skip-gram model is used for node embedding learning. For a node v4, the input layer is its one-hot encoding and the output layer is the one-hot prediction for all its K neighbor nodes (e.g. node v1 and node v10)