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.

Infinite motif stochastic blockmodel for role discovery in networks.

Infinite Hierarchical MMSB Model for Nested Communities/Groups in Social Networks

Low-rank persistent probability representation for higher-order role discovery

Dynamic Infinite Mixed-Membership Stochastic Blockmodel

A stochastic block Ising model for multi‐layer networks with inter‐layer dependence

motif2vec: Motif Aware Node Representation Learning for Heterogeneous Networks

Role detection in bicycle-sharing networks using multilayer stochastic block models

A Block-based Generative Model for Attributed Networks Embedding

Inferring Social Roles and Statuses in Social Networks

A stochastic block model for community detection in attributed networks

Scalable inference for stochastic block models

Alternative Blockmodelling

Probabilistic Community And Role Model For Social Networks

Stochastic Block Models with Multiple Continuous Attributes

Quantifying multiple social relationships based on a multiplex stochastic block model

Analyzing Online Transaction Networks with Network Motifs

Empirical Bayes estimation for the stochastic blockmodel

Self-supervised role learning for graph neural networks

Motif discovery in networks: A survey

Beyond Asymptotics: Practical Insights into Community Detection in Complex Networks