Representation learning using Attention Network and CNN for Heterogeneous networks

Highlights：

• A network embedding method for heterogeneous information network is proposed.

• Most network embedding methods require the use of meta-paths for semantic learning.

• The semantic information is learned by multi-typed edges without meta-paths here.

• The embeddings of all types of nodes in the network are learned at the same time.

• Our model performs better in node classification than most state-of-the-art methods.