Introduction to natural language processing

The course gives an overview of natural language processing (NLP) techniques applied (and applicable) to digital humanities issues. Symbolic methods are briefly discussed, but the course focuses primarily on recent machine learning methods (classifiers, neural networks, deep learning). The key elements to understand these methods are provided, as well as their application to various issues (corpus annotation, named entity recognition, opinion analysis). The programming language used is python, and students are frequently asked to dive into existing implementations (mainly Jupyter notebooks) and adapt them to their own data and research issues.

Teachers: 

• Carmen Brando
• Clément Plancq
• Thierry Poibeau

Practical information

• Public: Master's degree (possibly students beginning their thesis)
• Courses given in French
• Prerequisite: knowledge of Python programming
• 4 ECTS delivered
• 3 hours per week, over 8 weeks (from end of January to end of March)
• Course location: Ecole des Chartes, rue Richelieu, 75002 Paris
• Validation requirements: a short reflection text or an implementation linked to the course.