Graph Transformer for Graph-to-Sequence Learning
This addresses the limitation of graph neural networks in handling long-range dependencies for researchers in natural language processing, offering a novel method for graph representation learning.
The paper tackles the problem of graph-to-sequence learning by proposing a Graph Transformer model that allows direct communication between distant nodes for global graph structure modeling, achieving state-of-the-art results with improvements of up to 2.2 BLEU points in AMR-to-text generation and over 1 BLEU in syntax-based translation tasks.
The dominant graph-to-sequence transduction models employ graph neural networks for graph representation learning, where the structural information is reflected by the receptive field of neurons. Unlike graph neural networks that restrict the information exchange between immediate neighborhood, we propose a new model, known as Graph Transformer, that uses explicit relation encoding and allows direct communication between two distant nodes. It provides a more efficient way for global graph structure modeling. Experiments on the applications of text generation from Abstract Meaning Representation (AMR) and syntax-based neural machine translation show the superiority of our proposed model. Specifically, our model achieves 27.4 BLEU on LDC2015E86 and 29.7 BLEU on LDC2017T10 for AMR-to-text generation, outperforming the state-of-the-art results by up to 2.2 points. On the syntax-based translation tasks, our model establishes new single-model state-of-the-art BLEU scores, 21.3 for English-to-German and 14.1 for English-to-Czech, improving over the existing best results, including ensembles, by over 1 BLEU.