1. médialab Sciences Po
  2. News
  3. Linkage software for communications network analysis

Linkage software for communications network analysis

Pierre Latouche, Professor of Statistics at University Paris Descartes and Ecole Polytechnique, will present at the médialab seminar

Event, Research Seminar

Salle du médialab, 13 rue de l'Université, 75007 Paris

Tuesday may 21st, Pierre Latouche will present his work about "Le logiciel Linkage pour l’analyse des réseaux de communications. Etude de la recomposition politique en France via Twitter".

Biographie

Pierre Latouche is Professor of Statistics at University Paris Descartes and Ecole  Polytechnique, Paris, France. He received in 2011 the Ph.D. degree from  University Evry (France) for his work on network modeling and analysis.  He was Assistant Professor (2011-2017) and Associate Professor  (2017-2018) at University Paris 1 Panthéon-Sorbonne. His research  interests include network analysis, sparse inference, high-dimensional  data, graphical models, Bayesian analysis and variational approaches.  Pierre Latouche started his work on networks in  the mid 2000. He is interested in both methodological and theoretical  aspects. He developed the overlapping stochastic block model which  allows to look for overlapping clusters of nodes. More generally, over  the years, he has proposed many extensions to the original stochastic  block model. In particular, he is one the inventors of the linkage  methodology. Pierre Latouche is also part of the European Cooperation for Statistics of Network Data Science. 

Abstract

Due to the significant increase of communications between individuals  via social media (Facebook, Twitter, Linkedin) or electronic formats  (email, web, e-publication) in the past two decades, network analysis  has become a unavoidable discipline. Many random graph models have been  proposed to extract information from networks based on person-to-person  links only, without taking into account information on the contents.  This talk will introduce the stochastic topic block model (STBM), a probabilistic model for  networks with textual edges. We will address here the problem of  discovering meaningful clusters of vertices that are coherent from both  the network interactions and the text contents. A classification  variational expectation-maximization (C-VEM) algorithm will be proposed  to perform inference. Finally, we will rely on the methodology to study  the Enron political and financial scandals. We will also study the last  French presidential election. This work is implemented in the Linkage.fr  platform that I will present. 

Iframe https://medialab.sciencespo.fr/wp-content/uploads/2019/04/presentation-Linkage-PLatouche.pdf

Lecture recommandée

P. Latouche, C. Bouveyron, D. Marié, and G. Fouetillou. "Présidentielle 2017 : l’analyse des tweets renseigne sur les recompositions politiques". In : Statistique et Société 5.3 (2017).

Informations pratiques

Séminaire ouvert à tous, dans la limite des places disponibles.
Inscription préalable fortement recommandée: s’inscrire

Mardi 21 mai 2019 – 14h00 à 16h00
Salle du médialab, 13 rue de l’Université, 75007 Paris