Collective behavior in the big data era : can we enhance collective intelligence in human groups ?

lundi 07 mars 2016

7 mars 2016, 17h10

Collective behavior in the big data era : can we enhance collective intelligence in human groups ?

April 14 & 15, 2016 ; Toulouse School of Economics, Manufacture des Tabacs, Toulouse, France

Organizers : Adrien Blanchet (TSE), Matthieu Roy (LAAS), Clément Sire (LPT), Guy Theraulaz (CRCA)

Social and behavioral sciences have recently become “big-data sciences” mainly supported by the advances in digital and tracking technologies. These technologies have revolutionized the observation and quantification of social systems, improving to unprecedented levels the variety and precision of available data. Far from making theoretical modeling obsolete, these massively abundant data make a rigorous scientific method all the more necessary. Understanding how individuals interact with each other and give rise to complex collective behaviors requires not only statistical tools but also the development of mathematical models. Building predictive models of collective behavior become crucial as the concentration of populations in urban areas and the development of the Internet and mobile technologies may exacerbate the advent of dramatic events such as market panics or the propagation of rumors or false information. Such models, whose ingredients would be validated by experiments and data analysis, should not only help researchers to investigate of how complex processes are connected and how disruptions of these processes may contribute to the development of dramatic phenomena ; they could also ultimately open new perspectives to develop information systems to regulate the collective dynamics of human systems.

This workshop is intended to promote an interdisciplinary approach of collective behavior in human groups and to favor cross-disciplinary interactions between various communities : quantitative ethology, social sciences, economics, information technologies and statistical physics. It will mainly focus on the quantitative analysis and computational modeling of collective behavior in human groups. In particular we intend to discuss the following topics :

  • Under which conditions can interactions between individuals be modified to improve coordination or collaboration at a group level ?
  • Under which conditions can predictive models of decision-making be built from the computational analysis of digital traces ?
  • Can quantitative analysis of behavior reveal the information required to coordinate individuals' actions in a group or to help them to make efficient decisions ?

Workshop website : (Registration is free but mandatory if you want to participate)





10/04/2018 Cold-atom superradiance

John Martin (Université de Liège, Belgium) 10 avril 2018, 14h00   In my talk, I...