The successful candidate will be expected to lead the design and development of strategies for more transparent machine learning models to generate accurate cross-lingual representations for idiomatic language, as well as to contribute to the design and development of resources and evaluation of downstream tasks, like machine translation. For both lines of research, you will build on state-of-the-art approaches based on deep learning.

We are offering a Master's 2 internship in natural language processing on ‘Extracting knowledge about land use and land cover changes from textual data’. This internship will take place over a period of 6 months between January and June 2025 and will be co-supervised by CIRAD researchers from UMR TETIS as part of the TOSCA-CNES ARENA (Automatic Rule Extraction and Network Analysis) project.