Joël Legrand, MCF à Lille de l’équipe Inria Magnet, viendra nous présenter ses travaux le mardi 19 mars à 14h en salle A007 à CentraleSupélec (2 rue Edouard Belin, Metz).

Sa présentation est intitulée “Syntax-based Word Sequence Modeling using Deep Learning”.

Abstract:

Traditionally, natural language processing (NLP) relied on generative models with task specific and manually engineered features.
Recently deep learning approaches provided state of-the-art results in various fields such as computer vision, speech processing and natural language processing.
The central idea behind these approaches is to learn features and models simultaneously, in an end-to-end manner, and making as few assumptions as possible.
In NLP, word embeddings, mapping words in a dictionary on a continuous low-dimensional vector space, have proven to be very efficient for a large variety of tasks while requiring almost no a-priori linguistic assumptions.

In this talk, I will present some results on representations of segments of sentences using deep neural network models.
In particular, I will show how syntactic structures, such as consistuency and dependency syntactic trees, can be leveraged for the
purpose of solving NLP tasks that involve complex sentence-level relationships.
I will first introduce the key concepts of deep learning for NLP. I will then focus on two empirical studies concerning the tasks of
syntactic parsing using recursive neural networks (RNN) and relationship extraction from text using tree-LSTM with dependency structures.

Selected publications:
[1]. Joint RNN-based greedy parsing and word composition.
Joël Legrand and Ronan Collobert.
Proceeding of the 3rd International Conference on Learning
Representations (ICLR 2015)
http://arxiv.org/pdf/1412.7028

[2] Syntax-based Transfer Learning for the Task of Biomedical Relation Extraction
Joël Legrand, Yannick Toussaint, Chedy Raïssi, Adrien Coulet
Proceeding of the Ninth International Workshop on Health Text Mining and
Information Analysis (LOUHI 2018)
https://hal.inria.fr/hal-01869071/document

Short bio:
Joël Legrand received his MSc degree in Computer Science from the
Université de Lorraine and his Ph.D in Electrical Engineering from the
École Polytechnique Fédérale de Lausanne (EPFL, Switzerland).
He recently joined the  MAGNET team (Cristal, INRIA-CNRS-Université de
Lille) as an associate professor (maître de conférences).