Deep Kernel Network Embedding

This paper concerns the problem of network embedding (NE), whose aim is to learn a low-dimensional representation for each node in networks. We shed a new light to solve the sparsity problem where most of nodes including the new arrival nodes have little knowledge with respect to the network. A novel paradigm is proposed to integrate the multiple heterogeneous information from the subgraphs covering the target node instead of only the target node. Particularly, a probabiltiy distribution in subgraph space is contructed for each node, which is more effective to express the distinctive feature over the vertex domain compared to the traditional shallow representations. We boost NE performance by defining the convolution operation over the subgraph distributions that are efficient to evaluate and learn. Our method expliots the advantages of kernel method and deep learning such that the context semantics of subgraph distributions of nodes with dense links is transferred to the sparse nodes effectively via sharing model parameters. Experiments on four real-world network datasets demonstrate that our approach significantly outperforms state-of-the-art methods, especially on the representation learning for the nodes newly joining in the network.