PANTIN Jérémie

PhD student
Team : LFI
Arrival date : 10/01/2018
    Sorbonne Université - LIP6
    Boîte courrier 169
    Couloir 26-00, Étage 5, Bureau 504
    4 place Jussieu
    75252 PARIS CEDEX 05

Tel: +33 1 44 27 88 87, Jeremie.Pantin (at)

Supervision : Christophe MARSALA

Co-supervision : LESOT Marie-Jeanne

Model optimization for natural language processing tasks

There are many approaches to perform machine learning tasks in natural language processing. We focus on neural network models for abstractive text summarization task. Thus, the thesis focuses mainly on the optimization of these models and more particularly on model compression and distillation techniques as well as knowledge transfer. On one hand, we want to answer one of the problems of these models which is the computing power and, on the other hand, the need to have a large volume of data (compared to other approaches of machine learning). Fuzzy logic is intended to model logical reasoning on imprecise assertions. In our context we propose to use the theory of fuzzy sets as a basis for experimentation.