Ranking places in attributed temporal urban mobility networks

PLOS ONE RESEARCH ARTICLE Ranking places in attributed temporal urban mobility networks Mirco Nanni2☯, Leandro Tortosa1☯, José F. Vicent ID1☯*, Gevorg Yeghikyan3☯ 1 Department of Computer Science and Artificial Intelligence, University of Alicante, Alicante, Spain, 2 Institute of Information Science and Technologies, Italian National Research Council, Pisa, Italy, 3 Scuola Normale Superiore, University of Pisa, Pisa, Italy ☯ These authors contributed equally to this work. * a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Nanni M, Tortosa L, Vicent JF, Yeghikyan G (2020) Ranking places in attributed temporal urban mobility networks. PLoS ONE 15(10): e0239319. https://doi.org/10.1371/journal. pone.0239319 Editor: Alireza Abbasi, UNSW, AUSTRALIA Received: April 6, 2020 Accepted: September 4, 2020 Published: October 14, 2020 Copyright: © 2020 Nanni et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: We are willingly sharing the data starting from the APA values over time, along with all the code to compute those values, as well as the code for their subsequent spatio-temporal analysis. We also share the adjacency matrix and a the data vector. With this, the R code can be applied to obtain the centralities. Starting with the centralities, the rest are simulated with the supplied python function. We share an interactive website where one can try out different scenarios and see the results on the map. All this may be found in the following doi: https://doi.org/ 10.6084/m9.figshare.12720767.v1. Abstract Drawing on the recent advances in complex network theory, urban mobility flow patterns, typically encoded as origin-destination (OD) matrices, can be represented as weighted directed graphs, with nodes denoting city locations and weighted edges the number of trips between them. Such a graph can further be augmented by node attributes denoting the various socio-economic characteristics at a particular location in the city. In this paper, we study the spatio-temporal characteristics of “hotspots” of different types of socio-economic activities as characterized by recently developed attribute-augmented network centrality measures within the urban OD network. The workflow of the proposed paper comprises the construction of temporal OD networks using two custom data sets on urban mobility in Rome and London, the addition of socio-economic activity attributes to the OD network nodes, the computation of network centrality measures, the identification of “hotspots” and, finally, the visualization and analysis of measures of their spatio-temporal heterogeneity. Our results show structural similarities and distinctions between the spatial patterns of different types of human activity in the two cities. Our approach produces simple indicators thus opening up opportunities for practitioners to develop tools for real-time monitoring and visualization of interactions between mobility and economic activity in cities. Introduction The ever-growing availability of large scale data sources pertaining to human activities in contemporary cities and the fact that the socio-economic and technological systems lend themselves adequately to representation through discrete elements and interactions between them have led recent years to witness an unprecedented increase in modelling of such complex systems using network theory [1]. In urban science, there has been a significant research interest towards understanding urban systems particularly through modelling road structures, human mobility, traffic flow, and economic activity through a complex networks approach [2–4]. In such a setting, distinct elements in a city such as road junctions or neighbourhoods are typically represented as the PLOS ONE | https://doi.org/10.1371/journal.pone.0239319 October 14, 2020 1 / 25 PLOS ONE Funding: This work is supported by the Spanish Government, Ministerio de Economı́a y Competividad, grant number TIN2017-84821-P. It is also funded by the EU H2020 programme under Grant Agreement No. 780754, “Track & Know”. Competing interests: The authors have declared that no competing interests exist. Ranking places in attributed temporal urban mobility networks network nodes, while the heterogeneous connections or interactions between them, such as road segments, passenger flows, activity correlations represent the edges in the network [5, 6]. Further, depending on the focus of the research, various statistical and graph-theoretical properties of the network can be studied to gain valuable insights about the urban spatial, temporal and socio-economic structures. Following this approach, several studies have analysed mobile phone usage, taxi or private car GPS trajectories, smart card, geo-located social media, and classical census data for inferring systemic patterns both at the individual and aggregate level [7–11]. An area of research of particular interest in complex network theory is the study of the importance of nodes or edges in a network through centrality measures. Such measures are typically based on local and global network connectivity structures and include a variety of types: degree [12], closeness [13], betweenness [14], eigenvector [15], PageRank [16], etc. However, these conventional centrality metrics measure the importance of nodes by considering only the network topology regardless of the intrinsic information on these nodes such as their behaviour, type or some other, domain-specific attribute. Since many kinds of real-world networks call for such node attributes, several centrality measures have recently been proposed extending the widely used centrality measures to accommodate node attributes [17–19]. This becomes especially relevant in urban modelling, as locations in a city possess quantitative and qualitative characteristics irrespective of the connectivity structure of the network of interactions with other locations. Such characteristics may describe the availability and quantity of such urban features as parking lots, restaurants, real estate prices, population density, etc., qualitatively enhancing urban networks. Another important line of research in complex networks is temporal network theory: the study of the evolution and behaviour of networks over time. Temporal networks integrate network science with time-series analysis and contribute greatly to the modelling of epidemic spreading, transportation optimization, biological systems, as well as social networks [20]. Although some recent work has focused on analysing the spatial patterns of different urban features [5, 21], studying urban networks with centrality measures [17, 22, 23], as well as modelling the evolution of urban interaction networks over time [24], we still have a po (...truncated)