Open PhD student position: Analysing textual bias with explainable methods
[UPDATE: THE POSITION HAS BEEN FILLED]
Context
We are looking for a PhD candidate for the ANR project Slant (Spin and Bias in Language Analyzed in News and Texts) to work on text comparison and identifying language bias in text corpora. The successful candidate will integrate the Melodi team at the Institut de Recherche en Informatique (IRIT) in Toulouse, a team working on Natural Language Processing and Knowledge representation. The Slant project is a collaboration between the team Melodi, the team MAGNET at INRIA Lille and the ICRST of the University of Luxembourg.
Topic
The Slant project aims at identifying bias in text in general, in
order e.g. to understand differences in events coverage in media, by locating linguistic shifts from the lexical level to the rhetorical and argumentative ones. The generic approach aims at proposing automatic models identifying similarities and differences between texts on the same topic. The goal will be to also make explicit the information signaling these similarities or differences.
The thesis project will develop models to perform the comparison by enriching current models for semantic comparisons, — for now limited to sentence or to texts but without taking into account their structure — thus by considering argumentative and rhetorical dimensions.
Within Slant, a engineer has already been hired to collect news text to support the research project.
Profile
- The applicant must hold a Master in Computer Science, Natural Language Processing or Applied Mathematics.
- Knowledge or experience in one of the following fields: Machine Learning, Natural Language Processing.
Conditions/Environment
Location: team Melodi, laboratoire IRIT, Toulouse
Supervisors: Philippe Muller, Chloé Braud
PhD contract: 3 years
Starting: October 2020
Salary: 1875E / month (gross)
Applications should be sent to philippe.muller@irit.fr and chloe.braud@irit.fr, and should include: a CV detailing the formation and experience, especially in research ; a cover letter specifying your research interests ; Master grades ; if possible, references we could contact for recommandation.
We will examine applications as they arrive, and until the position is filled.
Project website: https://www.irit.fr/Slant/
Melodi team website: https://www.irit.fr/departement/intelligence-artificielle/equipe-melodi/