Authors: Mirco Nanni, Agnese Bonavita, Riccardo Guidotti

Exploratory: Sustainable Cities for Citizens

Classifying a geographical territory into semantic categories is one of the most common tasks in research areas such as urban geography, urban planning and mobility data analytics [1]. Characterizing human mobility is a key component of this process, and it is well known that mobility often does not work the same way across different regions. A movement pattern in a mountainous countryside may have other implications than the same pattern has in the suburbs of a large town. The movement trajectories in a planned city with rectangular streets and strict zoning laws might be completely different than the ones in a town that has grown organically without any clear structure. Therefore, any kind of property that was learned in a particular area, in general cannot simply be assumed to hold in another one. 

The paper presented in this post aims at making a first step towards the characterization of a geographical area. That is achieved through a range of quantitative measures that provide a multilayer description of urban regions and are a means for displaying differences between cities, municipalities, or other geographical units. Such a numerical description of urban areas can have a wide spectrum of applications. Among them, the measures presented in this work can be used as an input for geographical transfer learning, that is the transformation of knowledge gained in one geographical region in order to apply it to another region. This problem will be considered as a case study for the extracted indicators.

Authors consider two main approaches: (i) computing features that describe each area isolated from the others, that we call local city indicators; and (ii) computing features that describe its relation with the others, named global city indicators. 

The first group covers four different families of measures: spatial concentration indexes of human activities; network features of intra-city traffic flows; mobility characteristics of the individual mobility, obtained from networks that represent the places and movement of single

users; last, characteristics of road networks and how traffic is distributed in them. The group of global city indicators, instead, looks at the mobility between cities as a graph, where each city is represented by a node, and extracts network features for each node. 

Both the complete network and the ego-network for each city are considered.

After describing all the city indicators a mobility prediction problem is introduced, and authors use it to test how many predictive models are transferable across different regions. In particular, we study the relationship between transferability between two areas, i.e. the performances of a model built on one area and used to make predictions on the other one, and their similarity in terms of city indicators. The results confirm our hypothesis that cities with similar indicators are more likely to be transfer-compliant, this providing a first guide to understand which predictive models can be reused in other areas.

Finally, a key feature of this work is that all methods are implemented in a way that makes it possible to automatically calculate all characteristics for hundreds of different cities and entire regions. The resulting software (a Python library) enables the user to process an unlimited amount of data simply by passing a database with trajectories and a list containing the positions of the geographical areas of interest as an input.

The paper describing this research has been accepted to the workshop “BMDA 2021: Fourth International Workshop in Big Mobility Data Analytics”.

This work is partially supported by the European Community H2020 programme under the funding scheme Track &Know (Big Data for Mobility Tracking Knowledge Extraction in Urban Areas), G.A. 780754, https://trackandknowproject.eu/ and SoBig-Data++, G.A. 871042, http://www.sobigdata.eu.

Post written by: Agnese Bonavita

[1] Harold Carter. 1995. The Study of Urban Geography. E. Arnold publications.

[2] Gennady Andrienko et al. 2020. (So) Big Data and the transformation of the city. International Journal of Data Science and Analytics (2020).

[3] European Commission and European Investment Bank, 2016. “Smart Cities & Sustainable Development” Program in Europe.