arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2605.25548cs.LGcs.AI
收藏

Si'multaneous 'S'patial-'T'emporal Message Passing for Dynamic Graph Representation Learning

SiST-GNN: Simultaneous Spatial-Temporal Message Passing for Dynamic Graph Representation Learning

Department of Computer Science and Engineering(计算机科学与工程系);Indian Institute of Technology Gandhinagar(印度理工学院甘地纳加尔)

Shubhajit Roy, Anirban Dasgupta

AI总结:

提出SiST-GNN,通过在一个消息传递操作中融合空间和时间信号,实现动态图表示学习的联合推理,在链接预测任务上超越先前方法109%-277%。

摘要:

操作于快照序列的动态图神经网络(DGNN)通常分为两类:\emph{时间优先}方法先构建每个节点的时间嵌入,然后进行空间聚合;而\emph{空间优先}方法则颠倒这一顺序,将图卷积的输出馈送到下游时间模块。无论哪种情况,严格的顺序迫使第二阶段消耗第一阶段已压缩的摘要,排除了对拓扑和演化的联合推理;具体而言,消息传递算子永远无法根据邻居的\emph{过去}轨迹来加权其贡献。本文介绍了 extbf{SiST-GNN}( extbf{Si}multaneous extbf{S}patial- extbf{T}emporal extbf{GNN}),它在单个消息传递操作中融合两种信号,而不是将它们串联。具体地,在每个快照中,我们为每个节点维护一个循环隐藏状态来总结其历史,将其与节点当前特征向量配对,并将该配对视为由跨时间边连接的两个节点;在此时间增强图上运行标准图卷积,得到更新后的表示。我们的实证研究涵盖九个公开基线和十四个模型-数据集组合,覆盖固定分割和实时更新评估场景。在每个公开基准上,SiST-GNN在链接预测任务中相对于最强先前方法,在固定分割设置中提升109%-277%,在实时更新设置中提升68%-194%。我们还通过离散化底层连续时间事件流,构建了三个动态节点分类任务;在此,SiST-GNN以7%-22%的优势击败领先的离散时间(DTDG)基线,并与直接消费原始事件的连续时间(CTDG)方法相匹配。

英文摘要:

Dynamic graph neural networks (DGNNs) that operate on snapshot sequences typically fall into one of two categories. \emph{Temporal-first} approaches build per-node temporal embeddings and only afterward perform spatial aggregation, whereas \emph{Spatial-first} approaches invert this order, feeding the output of a graph convolution into a downstream temporal module. In either case, the rigid sequencing forces the second stage to consume an already-compressed summary produced by the first, ruling out joint reasoning over topology and evolution; effectively, the message-passing operator never gets to weight a neighbor's contribution by that neighbor's \emph{past} trajectory. This paper introduces \textbf{SiST-GNN} (\textbf{Si}multaneous \textbf{S}patial-\textbf{T}emporal \textbf{GNN}), which fuses the two signals inside a single message-passing operation rather than chaining them. At each snapshot, we maintain a recurrent hidden state per node that summarises its history, pairs it with the node's current feature vector, and treats the pair as two nodes joined by a cross-time edge; running a standard graph convolution on this temporally augmented graph yields the updated representation. We compare against fourteen link-prediction baselines under both the fixed-split and live-update evaluation regimes, and eleven baselines on node classification. Across the public benchmarks, SiST-GNN improves on the strongest prior method in link prediction by 1-18\% in the fixed-split setting, and is the leading learned method on five of six datasets in the live-update setting, improving on the strongest prior method by 1-158\% there. We additionally derive three dynamic node-classification tasks by discretizing the underlying continuous-time event streams; here SiST-GNN beats the leading discrete-time (DTDG) baseline by 7-23\% and matches continuous-time (CTDG) methods that consume the raw events directly.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑