PSL Week : NLP for Social Sciences

• Accueil
• Formation
• PSL Week : NLP for Social Sciences

The aim of the course is to provide theoretical and practical training in modern NLP methods for research in the social sciences (Economics, Management, Sociology, etc.). It is intended for students in the process of completing their thesis or for master's students who are considering pursuing a doctorate and adopting these rapidly developing methods. 

Dates & place:

21/25 November - 9h-12h : 14h-17h 

24 rue Saint-Georges 75009 Paris

Room I101 (Amphi Pierre Miquel) - from Monday to Thursday and room I102 on Friday

contact : christophe.benavent@dauphine.psl.eu

Public:

This course is primarily intended for master students at PSL University, but also for PhD students. Places are available for PhD students or colleagues from other universities (limited number, Prerequisite : Prior training in r and/or python is required.)

Application form is here

Syllabus:

• Session 1 : Tokenize and annotate
• Session 2 : Semantic spaces
• Session 3 : Topic modeling
• Session 4 : Embeddings
• Session 5 : ML for text
• Session 6 : introduction to Transformers
• Session 7, 8, 9 : hackathon with an Administration user experiences data set.
• Session 10 : students presentations

The lecture will be delivery principally in r, but the parallel procedure with python will be also studied. All materials (script, data and references) will be available at 

https://github.com/BenaventC/NLP_lecture_PSLWeek

r and Rstudio (with tidyverse) has to be installed before.

Team:

Christophe Benavent, Bruno Chaves Ferreira, José Carlos Romero Moreno and Olivier Caron from the Acss Institute.