In a metropolis, people movements design intricate patterns that change on very short temporal scales. Population mobility obviously is not random, but driven by the land uses of the city. Such an urban ecosystem can interestingly be explored by integrating the spatial analysis of land uses (through ecological indicators commonly used to characterize natural environments) with the temporal analysis of human mobility (reconstructed from anonymized mobile phone data). Considering the city of Milan (Italy) as a case study, here we aimed to identify the complex relations occurring between the land-use composition of its neighborhoods and the spatio-temporal patterns of occupation made by citizens. We generated two spatially explicit networks, one static and the other temporal, based on the analysis of land uses and mobile phone data, respectively. The comparison between the results of community detection performed on both networks revealed that neighborhoods that are similar in terms of land-use composition are not necessarily characterized by analogous temporal fluctuations of human activities. In particular, the historical concentric urban structure of Milan is still under play. Our big data driven approach to characterize urban diversity provides outcomes that could be important (i) to better understand how and when urban spaces are actually used, and (ii) to allow policy makers improving strategic development plans that account for the needs of metropolis-like permanently changing cities.

中文翻译：

---

米兰城市空间使用的时变网络

在大都市中，人们的动作设计出复杂的模式，这些模式会在非常短的时间尺度上变化。人口流动显然不是随机的，而是由城市的土地利用驱动的。通过将土地利用的空间分析（通过通常用于表征自然环境的生态指标）与人类流动性的时态分析（从匿名手机数据中重建）结合起来，可以探索这样的城市生态系统。以米兰市（意大利）为例，我们的目的是确定米兰市周边社区的土地利用结构与公民的时空居住格局之间的复杂关系。在分别分析土地使用和手机数据的基础上，我们生成了两个空间显式的网络，一个是静态的，另一个是时间的。在两个网络上进行的社区检测结果之间的比较表明，就土地用途组成而言相似的社区不一定具有人类活动类似的时间波动特征。尤其是，米兰历史悠久的同心城市结构仍在发挥作用。我们用大数据驱动的方法来表征城市多样性，所提供的结果可能很重要（i）更好地了解实际使用空间的方式和时间，以及（ii）允许政策制定者改善战略发展计划以解决大都市的需求，就像永久变化的城市。