We would like to invite you the CSMES Open House next Tuesday at 17h30 in Grenelle G44. This time the theme is Digital Humanities. There will be drinks and nibbles, and two very interesting speakers.
Julien Randon-Furling is a mathematician/physicist from the Sorbonne who works in complex systems, who will be talking about how ideas in opinion dynamics, language evolution, urban growth and social networks can be connected and understood. I include his full abstract below.
Dana Kianfar is a recent AUP graduate (CS major, math minor) now doing a PhD at the EPFL (Lausanne) . His talk is entitled "Inside Google Translate: Identifying Linguistic Capitalism and Algorithmic Bias On The Web". The full abstract is below as well.
Looking forward to seeing you there,
Julien Randon-Furling abstract:
Abstract: How can simple maths say something even vaguely relevant on things as varied and intrinsically complex as opinion dynamics, language evolution, urban growth, or ancient and contemporary social networks?!
Well, once upon a time, in the course of the nineteenth century to be more precise, theoretical physicists came to realize that Newton's equations ans laws, beautiful as they may be, did not seem to be the best mathematical tools if one wanted to describe a host of interacting particles... The likes of Maxwell and Boltzmann then imagined that such systems would be best approached by means of statistical tools. And that is how, a few decades before the quantum revolution, statistical physics brought probabilities at the heart of the "hard" sciences.
A hundred and fifty years on, these ideas have known a formidable development, the latest step being a tremendous interdisciplinary joint venture. Indeed, some of the key concepts (such as collective phenomena, tipping points and emergence) as well as the mathematical objects that physicists have tailored to study plasmas and spin glasses are now found to be extremely well suited to apprehend some of the most beautiful subtleties that human and social systems can exhibit.
In this talk, I will try to give an idea of the sort of simple maths and complex results one may encounter in the fast-developing science of complex systems, focusing in particular on some recent developments in the study of urban segregation dynamics."
Dana Kianfar abstract:
"We seek to identify and study pivotal features in text produced by algorithms. The rapid development of these algorithms cannot be viewed at independently from the economic context in which they operate. The widespread use of machine translation, autocomplete, text-completion and suggestion services mediate our textual expression, while adding more algorithmic text to the text corpora of the web, which other algorithms may use to model natural language. This project intends to progress in the development of methods and tools for monitoring this evolution, distinguishing algorithmic texts from texts produced by humans, and build computational models giving way to new hypotheses in order to understand this global linguistic evolution.