提出GelGT模型，提升关系图中长距离依赖建模能力
Gaussian Relational Graph Transformer
- 设计结构语义协同采样策略，保留关键连接并过滤无关信息
- 引入可学习高斯偏置的注意力机制，在子图上动态编码时间依赖
- 在真实数据集上性能领先，最高提升13.8%，适合时序关系建模任务
关系图学习将关系型数据库建模为图，在多种关系预测任务中表现优异。然而，现有方法因消息传递机制导致信息衰减，难以捕捉长距离依赖；近期的关系图变压器也未能有效联合建模结构、语义和时间信息。本文提出GelGT——一种高斯关系图变压器，显式解决上述挑战。GelGT采用结构-语义协同采样策略，在保持结构连通性的同时过滤无关语义信息；并在采样子图上引入带有可学习高斯偏置的高斯图注意力机制，动态编码时间依赖。在多个真实数据集上的大量实验表明，GelGT在下游任务中达到当前最优性能，预测准确率最高提升13.8%。
原文摘要 · Abstract (English)
Relational graph learning models relational databases as graphs and has demonstrated superior performance on a wide range of relational predictive tasks. However, existing methods struggle to capture long-range dependencies due to information decay in their message-passing mechanisms, and recent relational graph transformers remain limited in jointly modeling structural, semantic, and temporal information. In this paper, we propose GelGT, a Gaussian relational graph transformer that explicitly addresses these challenges. GelGT introduces a structure-semantic collaborative sampling strategy to preserve structural connectivity while filtering irrelevant semantic information, and incorporates a Gaussian graph attention mechanism with a learnable Gaussian bias on the sampled subgraphs to dynamically encode temporal dependencies. Extensive experiments on various real-world datasets demonstrate that GelGT achieves state-of-the-art downstream task performance, with up to a 13.8% improvement in predictive performance.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。