no code implementations • ACL (IWPT) 2021 • Xudong Zhang, Joseph Le Roux, Thierry Charnois
We review two features of mixture of experts (MoE) models which we call averaging and clustering effects in the context of graph-based dependency parsers learned in a supervised probabilistic framework.
no code implementations • 23 Oct 2023 • Francesco Demelas, Joseph Le Roux, Mathieu Lacroix, Axel Parmentier
Given any duals for these constraints, called Lagrangian Multipliers (LMs), it returns a bound on the optimal value of the MILP, and Lagrangian methods seek the LMs giving the best such bound.
1 code implementation • 22 Jun 2022 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
Starting from a deep probabilistic generative model with attention, we measure the interaction between latent variables and realizations of syntactic roles and show that it is possible to obtain, without supervision, representations of sentences where different syntactic roles correspond to clearly identified different latent variables.
1 code implementation • NAACL 2022 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
In the attention of Transformers, keys handle information selection while values specify what information is conveyed.
no code implementations • 21 Mar 2022 • Moussa Kamal Eddine, Nadi Tomeh, Nizar Habash, Joseph Le Roux, Michalis Vazirgiannis
Like most natural language understanding and generation tasks, state-of-the-art models for summarization are transformer-based sequence-to-sequence architectures that are pretrained on large corpora.
Abstractive Text Summarization Natural Language Understanding
1 code implementation • EMNLP (insights) 2021 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
We compare the simplified versions to standard SSVAEs on 4 text classification tasks.
1 code implementation • 24 Dec 2020 • Ghazi Felhi, Joseph Le Roux, Djamé Seddah
We present an unsupervised method to obtain disentangled representations of sentences that single out semantic content.
no code implementations • COLING 2020 • Yash Kankanampati, Joseph Le Roux, Nadi Tomeh, Dima Taji, Nizar Habash
In this paper we present a parsing model for projective dependency trees which takes advantage of the existence of complementary dependency annotations which is the case in Arabic, with the availability of CATiB and UD treebanks.
no code implementations • JEPTALNRECITAL 2020 • Davide Buscaldi, Ghazi Felhi, Dhaou Ghoul, Joseph Le Roux, Ga{\"e}l Lejeune, Xu-Dong Zhang
Dans notre travail nous nous sommes int{\'e}ress{\'e} {\`a} deux questions : celle du choix de la mesure du similarit{\'e} d{'}une part et celle du choix des op{\'e}randes sur lesquelles se porte la mesure de similarit{\'e}.
no code implementations • CONLL 2019 • Joseph Le Roux, Antoine Rozenknop, Mathieu Lacroix
We present a new method for transition-based parsing where a solution is a pair made of a dependency tree and a derivation graph describing the construction of the former.
no code implementations • JEPTALNRECITAL 2019 • Davide Buscaldi, Dhaou Ghoul, Joseph Le Roux, Ga{\"e}l Lejeune
Pour la ta{\^c}he d{'}indexation nous avons test{\'e} deux m{\'e}thodes, une fond{\'e}e sur l{'}appariemetn pr{\'e}alable des documents du jeu de tset avec les documents du jeu d{'}entra{\^\i}nement et une autre m{\'e}thode fond{\'e}e sur l{'}annotation terminologique.
no code implementations • JEPTALNRECITAL 2018 • Davide Buscaldi, Joseph Le Roux, Ga{\"e}l Lejeune
Notre premi{\`e}re m{\'e}thode est fond{\'e}e sur des lexiques (mots et emojis), les n-grammes de caract{\`e}res et un classificateur {\`a} vaste marge (ou SVM).
no code implementations • EMNLP 2017 • Caio Corro, Joseph Le Roux, Mathieu Lacroix
We present a new method for the joint task of tagging and non-projective dependency parsing.