Modelling human mobility is crucial in several areas, from urban planning to epidemic modeling, traffic forecasting, and what-if analysis. On the one hand, existing models focus mainly on reproducing the spatial and temporal dimensions of human mobility, while the social aspect, though it influences human movements significantly, is often neglected. On the other hand, those models that capture some social aspects of human mobility have trivial and unrealistic spatial and temporal mechanisms. In this paper, we propose STS-EPR, a modeling framework that embeds mechanisms to capture the spatial, temporal, and social aspects together. Our experiments show that STS-EPR outperforms existing spatial-temporal or social models on a set of standard mobility metrics and that it can be used with a limited amount of information without any significant loss of realism. STS-EPR, which is open-source and tested on open data, is a step towards the design of mechanistic models that can capture all the aspects of human mobility in a comprehensive way.

中文翻译：

---

考虑空间、时间和社会维度的人类流动建模

从城市规划到流行病建模、交通预测和假设分析，人类流动建模在多个领域都至关重要。一方面，现有模型主要关注再现人类流动的空间和时间维度，而社会方面虽然对人类运动有显着影响，但往往被忽视。另一方面，那些捕捉人类流动的某些社会方面的模型具有微不足道和不切实际的空间和时间机制。在本文中，我们提出了 STS-EPR，这是一个建模框架，它嵌入了将空间、时间和社会方面一起捕获的机制。我们的实验表明，STS-EPR 在一组标准移动性指标上优于现有的时空或社会模型，并且它可以在有限的信息量下使用，而不会显着降低真实性。STS-EPR 是开源的并在开放数据上进行测试，它是朝着机械模型设计迈出的一步，该模型可以全面捕捉人类移动的所有方面。