Abstract: Highlights•We conduct a comprehensive survey on deep graph representation learning algorithms.•We propose a new taxonomy of existing state-of-the-art literature.•This survey provides the practical and promising applications.•We state new perspectives and suggest challenging directions.
Loading