Watch Your Step: Learning Node Embeddings via Graph Attention

TL;DR

This paper introduces a trainable attention mechanism for graph node embeddings that replaces manual hyper-parameter tuning, leading to improved link prediction performance across diverse real-world datasets.

Contribution

We propose a novel attention model on the transition matrix power series that automatically learns hyper-parameters for graph embeddings, enhancing state-of-the-art results.

Findings

Improves link prediction accuracy by 20-45%.

Automatically learns graph-specific hyper-parameters.

Achieves better generalization to unseen data.

Abstract

Graph embedding methods represent nodes in a continuous vector space, preserving information from the graph (e.g. by sampling random walks). There are many hyper-parameters to these methods (such as random walk length) which have to be manually tuned for every graph. In this paper, we replace random walk hyper-parameters with trainable parameters that we automatically learn via backpropagation. In particular, we learn a novel attention model on the power series of the transition matrix, which guides the random walk to optimize an upstream objective. Unlike previous approaches to attention models, the method that we propose utilizes attention parameters exclusively on the data (e.g. on the random walk), and not used by the model for inference. We experiment on link prediction tasks, as we aim to produce embeddings that best-preserve the graph structure, generalizing to unseen information. We improve state-of-the-art on a comprehensive suite of real world datasets including social, collaboration, and biological networks. Adding attention to random walks can reduce the error by 20% to 45% on datasets we attempted. Further, our learned attention parameters are different for every graph, and our automatically-found values agree with the optimal choice of hyper-parameter if we manually tune existing methods.