A Self-Attention Network based Node Embedding Model

TL;DR

SANNE introduces a transformer-based self-attention network for inductive node embedding, effectively generating embeddings for unseen nodes and improving downstream tasks like node classification.

Contribution

It presents SANNE, a novel unsupervised model using self-attention for inductive node embedding, addressing the challenge of unseen nodes in graph learning.

Findings

Achieves state-of-the-art node classification results.

Effective in generating embeddings for unseen nodes.

Outperforms existing methods on benchmark datasets.

Abstract

Despite several signs of progress have been made recently, limited research has been conducted for an inductive setting where embeddings are required for newly unseen nodes -- a setting encountered commonly in practical applications of deep learning for graph networks. This significantly affects the performances of downstream tasks such as node classification, link prediction or community extraction. To this end, we propose SANNE -- a novel unsupervised embedding model -- whose central idea is to employ a transformer self-attention network to iteratively aggregate vector representations of nodes in random walks. Our SANNE aims to produce plausible embeddings not only for present nodes, but also for newly unseen nodes. Experimental results show that the proposed SANNE obtains state-of-the-art results for the node classification task on well-known benchmark datasets.