Introduction – A simulated training environment (‘sandbox’) for sharing questions and learning on the subject of local data
The profusion of local urban data and the deployment of increasingly sophisticated information and visualisation systems seem, at first glance, to be an appropriate response to the growing demand for information from private and public actors, political decision-makers, the general public and researchers. Rankings produced using synthetic indicators and complex dashboards, which are updated in real time and intended to enable a new governance of the smart city, mean we are now able to visualise and compare the internal differentiations of large cities, and examine different modes of urban organisation in relation to the formula- tions tested in the classic models developed by the Chicago and Los Angeles Schools (Judd & Simpson, 2011), and the various alternative reference frameworks.
However, these variegated big data streams also raise many conceptual and methodological questions around the manipulation of geographic infor- mation relating to space, time and thematics (Mathian & Sanders, 2014). In order to move beyond a rudimentary raw-data approach and offer stake-holders useful information in terms of territorial analysis, we must address two problematic assumptions:
- Interoperability of thematics: cross-tabulating indicators that are usually produced separately is complex because of the heterogeneous spatial definition of the geographical objects, which can lead to a lack of integrated, cross-cutting territorial governance.
- Spatial and temporal comparability at local level: the lack of inter-urban comparisons is often linked to a difficulty in harmonising data over space and time.
For the past three years, theCIST research group INFTER (Local Territorial Information) has been leading a reflection on these two topics within the framework of its Grandes métropoles project and the accompanying workshops . This reflection aims at strengthening the cross-cutting dimension of these problematics, notably through creating environments for shared learning and sharing methodologies. This document briefly outlines the main work carried out so far within this project and seeks to encourage other members of the CIST research federation to participate.
We begin by presenting the shared learning environments as well as the initial methodological challenges encountered during the construction of the database (1). We then address a number of fundamental aspects that are central to the comparative approach, albeit they are situated at different levels. First, from an inter-urban perspective, we discuss questions relating to data harmonisation using the example of land use and land cover (2). Second, drawing on demographic data, we examine the issues of comparative representation (3). Finally, we discuss how we assemble local territorial data that varies in granulometry, geometry and thematic, which is meaningful at the intra-urban level (Airbnb data, real estate prices data) (4).