Hierarchical Graph Capsule Network
This work aims to improve graph classification accuracy by better capturing hierarchical structures within graphs, which is a problem for researchers and practitioners working with graph-structured data.
This paper addresses the limitation of existing Graph Neural Networks (GNNs) in capturing hierarchical graph representations, which are crucial for graph classification. The authors propose the Hierarchical Graph Capsule Network (HGCN) to jointly learn node embeddings and extract graph hierarchies, achieving improved performance in experimental studies.
Graph Neural Networks (GNNs) draw their strength from explicitly modeling the topological information of structured data. However, existing GNNs suffer from limited capability in capturing the hierarchical graph representation which plays an important role in graph classification. In this paper, we innovatively propose hierarchical graph capsule network (HGCN) that can jointly learn node embeddings and extract graph hierarchies. Specifically, disentangled graph capsules are established by identifying heterogeneous factors underlying each node, such that their instantiation parameters represent different properties of the same entity. To learn the hierarchical representation, HGCN characterizes the part-whole relationship between lower-level capsules (part) and higher-level capsules (whole) by explicitly considering the structure information among the parts. Experimental studies demonstrate the effectiveness of HGCN and the contribution of each component.