发表机构
Xinjiang University（新疆大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出场景条件关系路由框架SCRR-Net，利用城市上下文联合控制空间依赖与跨任务知识迁移，在米兰和特伦托数据集上提升短信、流量和通话预测性能。
AI 中文摘要
城市蜂窝活动预测需要对异构时空信号进行联合建模，包括短信使用、移动网络流量和通话活动。现有方法通常将时间建模、空间关系学习和多信号预测分开处理，依赖固定的图结构或静态多任务学习方案，这限制了它们对不断变化的城市场景的适应性。我们提出SCRR-Net，一种场景条件空间关系路由框架，其中城市上下文信息联合控制空间依赖选择与跨任务知识迁移。SCRR-Net包含上下文编码器、空间图专家路由模块、时间Transformer编码器和任务知识路由模块。在米兰和特伦托数据集上的实验表明，SCRR-Net在短信、网络流量和通话活动预测方面持续优于竞争方法，同时提供可解释的路由行为。
英文摘要
Urban cellular activity forecasting requires jointly modeling heterogeneous spatiotemporal signals, including SMS usage, mobile network traffic, and call activity. Existing methods often separate temporal modeling, spatial relation learning, and multi-signal prediction, relying on fixed graph structures or static multi-task learning schemes, which limits their adaptability to changing urban scenes. We propose SCRR-Net, a scene-conditioned spatial relation routing framework in which urban contextual information jointly controls spatial dependency selection and cross-task knowledge transfer. SCRR-Net includes a context encoder, a spatial graph expert routing module, a temporal Transformer encoder, and a task knowledge routing module. Experiments on the Milano and Trento datasets demonstrate that SCRR-Net consistently outperforms competing methods on SMS, network traffic, and call activity forecasting, while providing interpretable routing behaviors.
CommentsAccepted in IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2026
