Abstract:Recent years have witnessed a surge of interest in machine learning on graphs and networks with applications ranging from vehicular network design to IoT traffic management to social network recommendations. Supervised machine learning tasks in networks such as node classification and link prediction require us to perform feature engineering that is known and agreed to be the key to success in applied machine learning. Research efforts dedicated to representation learning, especially representation learning using deep learning, has shown us ways to automatically learn relevant features from vast amounts of potentially noisy, raw data. However, most of the methods are not adequate to handle heterogeneous information networks which pretty much represents most real-world data today. The methods cannot preserve the structure and semantic of multiple types of nodes and links well enough, capture higher-order heterogeneous connectivity patterns, and ensure coverage of nodes for which representations are generated. We propose a novel efficient algorithm, motif2vec that learns node representations or embeddings for heterogeneous networks. Specifically, we leverage higher-order, recurring, and statistically significant network connectivity patterns in the form of motifs to transform the original graph to motif graph(s), conduct biased random walk to efficiently explore higher order neighborhoods, and then employ heterogeneous skip-gram model to generate the embeddings. Unlike previous efforts that uses different graph meta-structures to guide the random walk, we use graph motifs to transform the original network and preserve the heterogeneity. We evaluate the proposed algorithm on multiple real-world networks from diverse domains and against existing state-of-the-art methods on multi-class node classification and link prediction tasks, and demonstrate its consistent superiority over prior work.

Node Representation Learning in Graph via Node-to-Neighbourhood Mutual Information Maximization

Self-Supervised Node Representation Learning Via Node-to-Neighbourhood Alignment.

Graph Representation Learning Via Graphical Mutual Information Maximization.

motif2vec: Motif Aware Node Representation Learning for Heterogeneous Networks

Maximizing Mutual Information Across Feature and Topology Views for Learning Graph Representations

Heterogeneous Deep Graph Infomax

InfoMax Classification-Enhanced Learnable Network for Few-Shot Node Classification

Graph Representation Learning Beyond Node and Homophily

A 2 CMHNE

Unsupervised Hierarchical Graph Representation Learning by Mutual Information Maximization

Learning Network Representations with Neighboring Affinity

A2cmhne

Improving Node Representation by Boosting Target-Aware Contrastive Loss

Contrastive Representation Learning Based on Multiple Node-centered Subgraphs

Learning Node Representations from Noisy Graph Structures

Learning Representations by Graphical Mutual Information Estimation and Maximization

Representation Learning Based on Influence of Node for Multiplex Network.

Network2Vec: Learning Node Representation Based on Space Mapping in Networks

A Simplified Framework for Contrastive Learning for Node Representations

Robust Unsupervised Graph Representation Learning via Mutual Information Maximization

Geometric Graph Representation Learning via Maximizing Rate Reduction