Introduction aux méthodes statistiques en linguistique avec Python / Introduction to statistical methods in linguistics with Python
Lundi le 14 décembre 2015, grâce au soutien de l’Ambassade de France à Varsovie et de l’Institut français de Pologne à Varsovie, Alexandre Roulois (CNRS/Laboratoire de Linguistique Formelle (LLF)/Paris-Diderot) donne une conférence suivie d’un tutoriel : Introduction to statistical methods in linguistics with Python.
Organisateur : Wioletta Miskiewicz (CNRS/Archives Henri-Poincaré UMR 7117/e-LV), dans le cadre du programme « Humanités Numériques : projets, hommes, théorie 2 » réalisé en 2015/16 au Centre de la Civilisation Française de l’Université de Varsovie.
Statistics is a quantitative approach to research. Its relationship with math and science makes it at best ignored by the humanities community, at worst avoided. Indeed, humanities are known to attract students who flee the hard sciences. Even sadly, quantitative data are sometimes wrongly accused to destroy the « magic » in literary texts. On the contrary, I intend to demonstrate in my presentation that statistics can contribute to linguistics.
There is another false idea that grows in the mind of students in humanities I want to fight: programing is only for math people. As a matter of fact, programing has more to do with logic and linguistics than we would have thought. It is just like learning a new language and who is better at this than a student in humanities?
With only a few tools, we will build a program in Python that analyse a multilingual corpus of various texts (fiction, essay) and gives quantitative data and graphs as well.
Alexandre Roulois was at first chief assistant of the editorial board of a great search engine in France: « LeGuide.com ». But then he wanted to focus on his priorities and enrolled himself in a master’s degree in digital humanities with the objective to prepare the competitive exams of the public sector in the cultural field. This master’s degree was lead at the Centre for Advanced Studies in Renaissance (CESR/Tours).
In the early 21st, the CESR was really attractive because ahead of its time in the field of digital engineering with textual material, especially thanks to its flagship research program: the digital libraries of humanities. Through its operational know-how, the CESR was thus better able than any other to bring the skills necessary to manage the written heritage (production, distribution, conservation and promotion).
At the end of the training, he has immediately been received at a CNRS competition and was incorporated in the Institute of History and of Philosophy of Sciences and Techniques (IHPST), where he led projects in digital humanities and, among them, the e-LV publishing in TEI.
Many years later, he joins a research facility, the Laboratory of Formal Linguistics (LLF), where he can deepen his exploration of the digital tools to humanities.