Linkage software for the analysis of communication networks Go back to the news...

On may 21st 2019, Pierre Latouche will present his work at the médialab’s seminar.

The session will be about “Linkage software for the analysis of communication networks. Study of political recomposition in France via Twitter”


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.


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 platform that I will present.

Recommanded reading

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).