Vous êtes ici

Paraphrase generation / natural language generation

Lieu: 
Equipe: 
Contexte: 

IRISA (https://www.irisa.fr/) is the largest research laboratory dedicated to computer science in France, hosting more than 800 people and 40 research teams. Its activities spans all the fields of computer science. It is located in Rennes, Lannion, and Vannes.

The Expression team (https://www-expression.irisa.fr/) works on natural language processing (NLP), be it through texts, speech or gestures. In particular, it focuses on the expressive components of the human languages.

The opened position is part of the ANR TREMoLo project (http://tremolo.irisa.fr) hosted by the team and aimed at transforming the language register of texts, for instance mapping a text from the formal register to the casual one. This involves work on linguistic characterization, pattern mining and paraphrase generation. The activities are conducted on the French language.

 

Mission: 

The recruited person will work on a paraphase generation and propose solution to integrate register-specific stylistic constraints. She/he is expected to investigate on the use of statistical and neural paraphrasing systems,

  1. Training of a baseline systems using either statistical or neural approaches.

  2. Intregration of constraints formulated as sequential patterns.

  3. Organization of evaluation campaigns.

 

Profil / compétences: 
  • PhD in computer science

  • Recent skills in natural language processing

  • Recents skills in machine learning

  • Top academic and publication records

  • Good communication skills

  • Team work experience

  • Knowledge in French is a plus but is not required.

 

Diplôme requis: 
PhD
Lieu de travail: 
IRISA, Campus of Lannion
Type de contrat: 
CDD
Durée du contrat (en mois): 
24
Quotité: 
100%
Salaire Brut / Mens €: 
Competitive, depending on the experience
Date prévisionnelle d'embauche: 
As soon as possible
Candidater: 

Please send a cover letter and resume to :

Application deadline: until the post is filled.