Post-doctoral position (M/F) Identification of gendered expressions by vector representations on an...

The person recruited (M/F) will be in charge of developing unsupervised or semi-supervised natural language processing (NLP) techniques applied to corpora of automatic speech recognition outputs, to identify "gendered expressions" such as references to cultural stereotypes based on gender, traditional named entities or any reference to private life, age, physique, sexuality, skills, etc.
Secondarily, the analysis of biases in language models can also be conducted.

The corpora are made available by the project leader (Institut National de l'Audiovisuel) and are composed of: radio morning shows and television newscasts from the GMMP (Global Monitoring Media Project) corpus, French radio programs (culinary, economic, sports, and talk shows) for the study of incivilities (interruptions, insults, etc.), and reality TV shows (Loft Story 2001, Les Marseillais à Dubaï 2021). No annotation is available around gendered expressions. The person recruited will therefore have to use unsupervised or semi-supervised methods.

This work will be supervised by Dr. Sahar Ghannay (assistant professor in informatics) and Dr. Grouin (research engineer in informatics). The contract is funded by the French National Research Agency (ANR GEM 2019). The project is led by Dr. David Doukhan (French National Audiovisual Institute).

The GEM (Gender Equality Monitor) project aims at analyzing the interactions between women and men in the media (radio and television), and more particularly the differences in representations according to whether the person expressing him/herself is a woman or a man, according to his/her role (anonymous, journalist, politician, etc.), and according to the themes addressed. In this inter-disciplinary project, the computer science partners (including LISN) have the task of implementing the descriptors that will allow the humanities and social sciences partners to quantify and qualify the differences in representation. https://anr.fr/Projet-ANR-19-CE38-0012

The Laboratoire Interdisciplinaire des Sciences du Numériques (LISN) is an academic research lab located on the Saclay plateau which has been created in 2021 from the merger of the LIMSI and LRI laboratories. The research carried out at LISN covers a broad scientific spectrum and is internationally recognized.

The laboratory has more than 380 members in 16 research teams and 6 support services. The lab is entirely in a restricted research area (ZRR) which imply and administrative processing of all application forms..

The person recruited will work within the ILES team, in close collaboration with the researchers of the ILES and TLP teams involved in the project, within the Languages Sciences and Technology department.