MPool: Motif-Based Graph Pooling

Recently, Graph Neural Networks (GNNs) have emerged as a powerful technique for various graph-related tasks. Current GNN models apply different graph pooling methods that reduce the number of nodes and edges to learn higher-order structure in hierarchical way. However, these primarily rely on one-hop neighborhood do not consider graph. To address this issue, work, we propose multi-channel Motif-based Pooling method named (MPool) captures with motif also considers local global through combination selection clustering-based operations. In first channel, develop node selection-based by designing ranking model considering adjacency nodes. second cluster-based spectral clustering using adjacency. Finally, result each channel is aggregated into final representation. We perform extensive experiments demonstrate our proposed outperforms baseline classification tasks eight benchmark datasets.