arXiv cs.AI· Ahmet T\"uzen, Helge Langseth, Kjetil N{\o}rv{\aa}g·· 9 天前AI 评分22
Schema：通过软社区结构实现可扩展的层次化图生成
Scalable Hierarchical Graph Generation via Soft Community Structure
AI 导读
Schema 通过将参考图递归分解为软社区层次结构，把图生成拆分为节点属性合成、社区内边生成和跨社区桥接节点建模三个独立训练阶段，各阶段均不构建完整邻接矩阵。在四个真实属性图上，Schema 比任何其他生成属性的模型都更准确地恢复了局部与长程结构的平衡，同时仅复现少量参考边，并保持参考图的下游准确率。研究还在六个额外图上验证了可扩展性，最大规模达 1000 万节点。
正文 · 原文
Abstract:Generating large attributed graphs requires reproducing the topology, generating attributes jointly with the structure, and remaining scalable. Many real-world graphs exist as a single large graph, so a generative model has to generalize from the one graph it is fit on, without independent samples. We present Schema, which recursively decomposes a reference graph into a hierarchy of soft communities, assigning each node a membership distribution. Generation is then split into three stages, each trained independently: (1) synthesizing node attributes conditioned on soft memberships, (2) generating intra-community edges from local structural context, and (3) modeling inter-community connections over bridge nodes whose membership mass is distributed across several communities. No stage forms the full adjacency matrix, and each stage operates on a subgraph bounded by the community size. We also introduce an evaluation protocol that covers structural fidelity, memorization, downstream utility, and scalability. On four real-world attributed graphs, Schema recovers the balance between local and long-range structure more closely than any other model that generates attributes, while reproducing only a small fraction of the reference edges. It retains the downstream accuracy of the reference graph without raising it artificially above that level. Baselines that match its structural fidelity memorize the reference, while those with higher downstream accuracy either exceed the reference accuracy or fail to complete on the larger graphs. We measure scalability on six additional graphs with up to 10 million nodes.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.12163 [cs.LG] |
| (or arXiv:2610.12163v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.12163 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ahmet Tüzen [view email]
[v1]
Thu, 8 Oct 2026 15:39:53 UTC (819 KB)
来源：arXiv cs.AI · arxiv.org