Keynote: Natural Language Processing without supervision

In this talk Sebastian Riedel will give an overview of the work done at FAIR in the area of Natural Language Processing (NLP).

In particular, he will focus on FAIR’s work on unsupervised NLP with the goal of dramatically reducing the cost of NLP on new domains and languages.

He will discuss how we can train neural translation, style transfer and question answering models without any supervision at all.