Page de Pablo Jensen


Modeling social systems using big data

Everything changes, yet change is ill understood. The usual strategy consists in finding stable entities that allow to tame change by reducing it to quantitative evolutions. Qualitative transformations of matter - liquid water disappearing into the air or solidifying as ice – are translated by physics into changes in the configuration of stable entities, molecules or atoms. Social evolutions can sometimes be described as a simple, quantitative, variation in the sizes of different sub-populations: Population aging measured by the increase of the proportion of people older than 65 vs people younger than 25; higher proportion of biologists now (respect physicists) than at the beginning of the 20th century... This simple strategy is not always feasible, when transformations “break” the atoms or the categories themselves evolve. When biophysicists emerge in the 20th century, they cannot be counted as physicists or biologists: A new category has to be created. In ongoing work, we acknowledge the fact that many (the most important?) social transformations cannot be simplified using some “substance”, literally, what lies below, giving stability and continuity, no social atoms or stable categories. The challenge for the description of social dynamics is then : How to create relevant “evolving categories”, coherent flows that capture the essential processes at work? To make sense of a sequence of events (publication of scientific articles...), one cannot stick to a succession of instantaneous categories that would be adapted to each specific time. But on which grounds can we attribute a unity to a specific succession of events, without the help of an underlying substance ? Our method is based on the idea that the unity of an evolving social process rests on the continuity by which one transformation proceeds from another along an uninterrupted succession (Elias). In practice, our method combines two steps: first, a tentative instantaneous structure is independently created for each time step, and, second, the requirement of continuity allows to distinguish real long-term trends from artefactual fluctuations induced by the fuzziness in the (instantaneous) community definition.

Identifying global bridges in networks (paper accepted in the Journal of Complex Networks)

We have created a new indicator to identify global bridges in networks. The identification of nodes occupying important positions in a network structure is crucial for the understanding of the associated real-world system. Usually, betweenness centrality is used to evaluate a node capacity to connect different graph regions. However, we argue here that this measure is not adapted for that task, as it gives equal weight to “local” centers (\emph{i.e.} nodes of high degree central to a single region) and to “global” bridges, which connect different communities. This distinction is important as the roles of such nodes are different in terms of the local and global organisation of the network structure. In this paper we propose a decomposition of betweenness centrality into two terms, one highlighting the local contributions and the other the global ones. We call the latter \emph{bridgeness} centrality and show that it is capable to specifically spot out global bridges. In addition, we introduce an effective algorithmic implementation of this measure and demonstrate its capability to identify global bridges in air transportation and scientific collaboration networks.

Understanding the emergence of social structures using original sociological approaches

We work on the crucial question of the relations between collectives and individuals, the emergence of social “structures” in collaboration with sociologists MediaLab Sciences Po in Paris directed by Bruno Latour. This ongoing work includes both the development of original models that respect the richness of humans, who somehow simplify themselves by interacting and building collectives, institutions, while physicists have rather the opposite reflex of starting from “atomic” invididuals (as simple as possible) to bring out complex structures at the "macroscopic" level. A paper has appeared in the British Journal of Sociology (2012).

A more recent paper (published in JASSS ) proposes to shift from simulations of 'social atoms' to formal descriptions of social data, to improve the empirical description of society **************************************

Understanding the field of Education

La recherche en éducation couvre un territoire de divers sujets ce qui présente un défi d’intégration. Nous avons effectué une analyse bibliographique de la base de données internationale Scopus afin de décrire la recherche en éducation Voir l'article en français : paper

Understanding science dynamics using scientometrics

What is the science of complex systems? To obtain an empirical answer to that question, we have used scientometrics’ data from Web of Science. We have identified about 200 000 items which we then analyzed and visualized. We wanted these data to answer sociological and historical issues: how do different subdomains emerge? How experimental and theoretical disciplines can cooperate? Is there a coherence of the field "complex systems"? This work has been published by Journal of the American Society for Information Science and Technology (JASIST).
A more epistemological reflection on the theme of complexity, discussing the one proposed by Edgar Morin, was conducted within IXXI and resulted in a recent article in the journal Hermès .
We are currently applying this scientometric approach to scientific institutions, including ENS de Lyon. We prepared a set of maps of the ENS that are currently visible in its buildings to elicit feedback from researchers. This work was published in 2011 in the journal "Scientometrics" .

Understanding the uses of bycicle sharing systems

We study the bicycle sharing system from Lyon (Velov) with a group of researchers form the Complex Systems Institute, Rhône-Alpes ( and the Transport Economics Laboratory ( We have obtained the distances and journey times of all Vélo'v users over the 25th May 2005 - 14th December 2007 period (data provided by JC Decaux). From the aggregated data, we have shown that cyclists ride at an average of 13.5 km/h. However, as they take shortcuts, they really reach a speed equivalent to 15km/h relative to the circuits followed by motor vehicles. In contrast, cars move at average speeds between 10 and 15 km/h in European towns. There is better to come - the fastest 10% of cyclists reach 20km/h during the morning rush hour, overtaking the vehicles trapped in traffic jams. Pablo Jensen, Nicolas Ovtracht, Jean-Baptiste Rouquier and Céline Robardet, Note published in Transportation Research D (2010)

NB : Vous pouvez télécharger des articles de ce site à condition de respecter les règles du copyright.
NB : Please respect the copyright rules when accessing full lenght articles on this site.

Courriel: pablo.jensen at
Téléphone: +33 4 26 23 38 01