SCGFM-ART: Amortized Relational Transport for Structure-Centric Graph Foundation Models
作者: Xiaodong He, Xincheng Wang, Zhao Kang
分类: cs.LG, cs.AI
发布日期: 2026-09-17
备注: 21 pages, 6 figures
💡 一句话要点
提出SCGFM-ART以解决异构图领域的统一表示问题
🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 图基础模型 异构图 关系图谱 传输计划 迁移学习 结构对齐 图表示学习
📋 核心要点
- 现有的图基础模型在处理拓扑、图规模和特征语义的严重异构性时面临挑战,难以构建统一的表示空间。
- SCGFM-ART通过结构中心的框架和Amortized Relational Transport方法,将任意图对齐到共享的关系图谱,简化了图到基准的传输过程。
- 在14个跨域图和节点级分类任务中,SCGFM-ART实现了最先进的转移能力,平均排名分别为2.29和1.14,显著提升了推理速度。
📝 摘要(中文)
图基础模型(GFMs)旨在学习跨越严重异构图域的可转移表示。然而,拓扑、图规模和特征语义的严重领域转变阻碍了统一、领域无关的表示空间的构建。为此,本文提出了SCGFM-ART,一个结构中心的GFM框架,通过 amortized relational transport (ART) 将任意图对齐到共享的关系图谱上。关系图谱作为由有限的关系基准定义的通用坐标系统,而ART直接预测可重用的图到基准的传输计划,避免了耗时的Gromov-Wasserstein优化。在此框架下,SCGFM-ART将图分解为统一的表示,解决了结构和语义异构性。
🔬 方法详解
问题定义:本文旨在解决异构图域之间的统一表示问题,现有方法在拓扑和特征语义的转变下难以有效对齐图结构。
核心思路:SCGFM-ART通过构建一个共享的关系图谱,利用Amortized Relational Transport (ART) 直接预测图到基准的传输计划,从而避免了复杂的优化过程。
技术框架:该方法的整体架构包括关系图谱的构建、图的分解与对齐、以及基于ART的传输计划预测,形成了一个高效的图表示学习流程。
关键创新:最重要的创新在于提出了ART方法,能够在不进行迭代对齐的情况下,快速生成图到基准的传输计划,显著提高了推理效率。
关键设计:在设计中,模型通过有限测量的关系空间来严格建模图和基准,采用了特定的损失函数来保证学习目标的紧密近似理想关系覆盖。具体的参数设置和网络结构细节在实验部分进行了详细描述。
🖼️ 关键图片
📊 实验亮点
在14个跨域图和节点级分类任务中,SCGFM-ART取得了最先进的转移能力,平均排名分别为2.29和1.14。此外,实验证明其推理速度比传统方法快44.2到85.1倍,显示出显著的效率提升。
🎯 应用场景
SCGFM-ART的研究成果在多个领域具有潜在应用价值,尤其是在社交网络分析、生物信息学和推荐系统等需要处理异构图数据的场景中。通过提供高效的图表示学习方法,能够提升模型在不同领域间的迁移能力,推动相关技术的发展。
📄 摘要(原文)
Graph foundation models (GFMs) aim to learn transferable representations across severely heterogeneous graph domains. However, severe domain shifts in topology, graph scale, and feature semantics impede the construction of a unified, domain-agnostic representation space. To address this, we propose SCGFM-ART, a structure-centric GFM framework that aligns arbitrary graphs onto a shared relational atlas via Amortized Relational Transport (ART). The relational atlas serves as a universal coordinate system defined by a finite set of relational landmarks (bases), while ART directly predicts reusable, end-to-end graph-to-base transport plans, bypassing costly runtime Gromov-Wasserstein optimizations. Under this formulation, SCGFM-ART decomposes a graph into a unified representation: globally via its relational response coordinates relative to the atlas, and locally via its node-to-role structural correspondences. These correspondences project disparate node attributes into a canonical role space, resolving structural and semantic heterogeneity within a singular alignment interface. Rigorously modeling graphs and atlas bases as finite measured relational spaces, we establish coordinate fidelity bounds, prove stability under predicted transport plans, and derive an amortized coverage bound that guarantees our learning objective tightly surrogates ideal relational coverage. Benchmarked across 14 cross-domain graph- and node-level classification tasks, SCGFM-ART achieves state-of-the-art transferability, securing superior average ranks of 2.29 and 1.14, respectively. Topological perturbation analyses demonstrate that node-role transport retains fine-grained structural nuances beyond global coordinates. On real-world benchmarks, the amortized formulation yields 44.2 to 85.1 times faster frozen target-domain inference by avoiding iterative alignment at test time.