Section: Overall Objectives

Overall Objectives

The past decade have witnessed an explosion of the amount of data harvested through digital systems and produced by human activity or from the large set of environmental sensors (IoT). These data are collected, analyzed, correlated and transformed to enable innovative services, which have strong, often disruptive, impact on societies. The datasphere is the new space resulting from these data, considered as a whole, independently of their control. It constitutes a dynamic complex system, much like the hydrosphere, where the basic constituents are bits of data in place of water molecules. Although the date of the inception of the datasphere can be debated, it is really at the turn of the century that its role became dominant and challenged the legacy organisation of societies.

The objective of the Datasphere team is to study the transformation of socio-economic and socio-ecological systems triggered by the diffusion of digital services. We propose a holistic view of the datasphere to apprehend global changes at a planetary scale, with a long term perspective, sometimes with a teleologic vision to understand the phenomena at play and model the interactions of the future. We also consider the digital transformation of socio-economic systems in relation with the challenges that the threats on the natural ecosystem of our planet impose on human societies. Both transformations happen contemporarily, and share parallel impact on the tension between local and global, vertical and horizontal.

A major goal of the project is to develop tools and methodologies in order to observe and analyse the ongoing changes induced by digital transformations. These tools are generally software systems that have to process large volume of heterogeneous data in order to harvest relevant metrics. For this purpose we are pursuing big data processing, machine learning, data visualization, cartography and graph analysis methods that are applicable to our specific needs and can be used in broader scopes.

From a methodological point of view, we aim (i) at interdisciplinary research with all relevant disciplines, and in particular social sciences, and (ii) when possible, analysis of large datasets, such as those from network activities, to investigate quantitatively global phenomena. The first aim raises classical difficulties of interdisciplinary research, but is carried on in a very favorable environment, namely the complex system institute, IXXI. For the second aim, we need to develop original data analysis techniques, new metrics on data flows related to social activities, as well as new visualisation methods to show the interdependencies between entities, from States to people and devices.