Deep Inductive Graph Representation Learning

This paper presents a general inductive graph representation learning framework called <tex-math notation="LaTeX">DeepGL</tex-math> for learning deep node and edge features that generalize across-networks. In particular, <tex-math notation="LaTeX">DeepGL</tex-math> begins by deriving a set of base features from the graph (e.g., graphlet features) and automatically learns a multi-layered hierarchical graph representation where each successive layer leverages the output from the previous layer to learn features of a higher-order. Contrary to previous work, <tex-math notation="LaTeX">DeepGL</tex-math> learns relational functions (each representing a feature) that naturally generalize across-networks and are therefore useful for graph-based transfer learning tasks. Moreover, <tex-math notation="LaTeX">DeepGL</tex-math> naturally supports attributed graphs, learns interpretable inductive graph representations, and is space-efficient (by learning sparse feature vectors). In addition, <tex-math notation="LaTeX">DeepGL</tex-math> is expressive, flexible with many interchangeable components, efficient with a time complexity of <tex-math notation="LaTeX">O(|E|)</tex-math> , and scalable for large networks via an efficient parallel implementation. Compared with recent methods, <tex-math notation="LaTeX">DeepGL</tex-math> is (1) effective for across-network transfer learning tasks and large (attributed) graphs, (2) space-efficient requiring up to 6x less memory, (3) fast with up to 106x speedup in runtime performance, and (4) accurate with an average improvement in AUC of 20 percent or more on many learning tasks and across a wide variety of networks.