DigiCosme Laboratory of Excellence opens
To mark the inaugural conference of the DigiCosme Laboratory of Excellence (Labex) on 12 and 13 September 2012, we take a look back at the origins and ambitions of this project in an interview with the lab's coordinator, Christine Paulin-Mohring.
The DigiCosme* Labex did not just materialise out of thin air ...
Since the 2000s, various changes have taken place that have improved the structure of computer science research and training activities at the Plateau de Saclay.
Inria's arrival was an initial turning point, promoting cooperation between the École Centrale, the École Polytechnique, ENS Cachan and Université Paris-Sud through joint project teams. This resulted in the Joint Computer Science Research Cluster (PCRI). These Inria projects were unique in their long-term approach. The researchers saw each other regularly, working together and taking advantage of their proximity to each other, over long periods (around ten years on average). Then came the Digiteo stage, with an extension of collaborations to other partners, such as the CEA List institute. The Labex should allow the computer science structure to be strengthened at a time when the laboratories involved are coming closer together geographically. It comprises 300 researchers and the same number of PhD students, spread among the sites of 11 institutions: the French atomic energy commission (CEA), the French national centre for scientific research (CNRS), the École Polytechnique, Supélec, Inria, Université Paris-Sud, the École Centrale Paris, ENS Cachan, ENSTA ParisTech, the Institut Mines-Télécom, and, finally, Université de Versailles St Quentin.
In principle, a Labex has to be multidisciplinary, yet you have chosen to focus on computer science ...
It's true that an interdisciplinary structure was one of the criteria, but evidently not the most decisive one, as our selection shows. By choosing to create a single-discipline Labex, we are not giving up on interdisciplinary collaboration. As part of the Idex (initiative of excellence), teams can continue to pursue cross-disciplinary research with biologists, physicists, etc. We already have fruitful relations with physics labs enabling us to observe computing systems as physical systems.
However, unlike other disciplines such as physics or mathematics, information and communication sciences are still a young discipline. They have to find their place and be recognised as a science in their own right, even if they do play an important role in helping other sciences. Computer scientists are most often drawn towards interfaces with other disciplines, but they must promote recognition of their own core discipline. We are witnessing explosive growth in the information society—that is clear—but if we want to reap the benefits, we must think deeply and come up with new models. So what we wanted to do with this Labex was defend the core of information sciences.
We do not seek to cover every issue. Rather, it is about working on fundamental areas, such as programming, data and communications, that lie at the core of our discipline. We have tried to identify challenges in each of these areas. At present, the main challenge is in relation to data, which are increasingly numerous, heterogeneous and distributed. We have also moved from the era of stand-alone computers to that of decentralised systems. The models change and the algorithms are modified as a result.
Does this return to the core mean you will be working solely with computer scientists? Or will you call upon other skills?
In fact, we will essentially be working on projects of a limited scale, as that ensures maximum researcher involvement. From a university point of view, the information and communication aspects of networks are not covered by the same sections or even the same laboratories. The Labex strengthens the synergies between these two fields, which lie at the heart of modern-day information systems. However, we also want to draw on skills from other disciplines. Digital systems have grown in size so much that the observation and modelling techniques used with them are getter closer to physics. Collaborations with physicists are therefore an obvious choice. We also want to develop links with social sciences.
You are focusing on the core of computer science: does that mean you will be preferring theoretical research to the applied variety?
I wouldn't make that distinction. In computer science, there is very little distance between theoretical and applied research. All the themes we have proposed are directly linked to applications. Certification and correction problems, for example, range from sensors to the most elaborate abstract models and are based on advanced logical and algorithmic results.
* The project was endorsed by the French national research agency (ANR) under the name DigiWorlds. Due to a conflict with a registered trademark, this name was abandoned in September 2012 and the Labex was renamed 'DigiCosme'.
An expert in multi-institutional work
After studying at the ENS de Jeunes Filles, which she entered in 1982, Christine Paulin-Mohring submitted a thesis, prepared in partnership with Inria and the ENS, at Université Paris 7 in 1989. 'I have always worked in a multi-institutional environment! ' As a research scientist at the CNRS and then a computer science professor at Université Paris-Sud in 1997, she took part in the first joint projects between laboratories. With her former PhD supervisor, she set up the first joint team between Inria and the LIP, a CNRS laboratory at ENS-Lyon. This team would become one of the first joint projects under the Inria Futurs programme, which would also involve Paris-Sud, the CNRS and the Ecole Polytechnique.
Excerpt from the interview with Christine Paulin-Mohring concerning 'a Labex for a discipline without boundaries: computer science', published on the Paris-Saclay Media site