GraphFormers: GNN-nested Transformers for Representation Learning on Textual Graph

TL;DR

GraphFormers introduces a novel architecture that integrates GNN components within transformer-based language models to improve textual graph representation learning by fusing text encoding and graph aggregation in an iterative process.

Contribution

It proposes a nested GNN-transformer architecture and a progressive training strategy, enabling more accurate and global understanding of node semantics in textual graphs.

Findings

Outperforms state-of-the-art baselines on three large-scale datasets

Maintains comparable efficiency with existing methods

Enhances the integration of textual and graph information

Abstract

The representation learning on textual graph is to generate low-dimensional embeddings for the nodes based on the individual textual features and the neighbourhood information. Recent breakthroughs on pretrained language models and graph neural networks push forward the development of corresponding techniques. The existing works mainly rely on the cascaded model architecture: the textual features of nodes are independently encoded by language models at first; the textual embeddings are aggregated by graph neural networks afterwards. However, the above architecture is limited due to the independent modeling of textual features. In this work, we propose GraphFormers, where layerwise GNN components are nested alongside the transformer blocks of language models. With the proposed architecture, the text encoding and the graph aggregation are fused into an iterative workflow, {making} each node's semantic accurately comprehended from the global perspective. In addition, a {progressive} learning strategy is introduced, where the model is successively trained on manipulated data and original data to reinforce its capability of integrating information on graph. Extensive evaluations are conducted on three large-scale benchmark datasets, where GraphFormers outperform the SOTA baselines with comparable running efficiency.