(Deadline extension)

UCLouvain is looking for:

a postdoctoral researcher in machine learning / natural language processing

- Full-time (100%) fixed-term contract of two years - for the Centre de traitement automatique du langage (Cental) within the Institut Langage & Communication (IL&C) in UCLouvain (Louvain-la-Neuve) - Start date : as soon as possible

This postdoctoral position offer is part of a research project led by the Cental (https://uclouvain.be/fr/instituts-recherche/ilc/cental) around legal data processing.

Regarding the concrete application, the project aims at automatizing the analysis of documents related to clinic trials (meeting minutes, legal documents, contracts, ...) to assess their compliance to RGPD. The proposed solution should thus be flexible enough to, on one hand, ensure that the model(s) can be adapted to the various document types and, on the other hand, limit the need of specialists' expertise for training data annotation. In consequence, the scientific core of this project is directly related fo the question of few-shot learning, which we intend to address through active learning and meta-learning.

The role of the hired postdoc will be to (1) develop the resources needed for learning, (2) implement an architecture that incorporates active learning and meta-learning, (3) evaluate the models and (4) implement the components into a web service. The postdoc will also be required to disseminate the results through scientific publications and/or reports.

Work environment:

CENTAL is part of the Institut Langage & Communication ( https://uclouvain.be/fr/instituts-recherche/ilc), in UCLouvain. This university is located in Louvain-la-Neuve, Belgium ( https://uclouvain.be/fr/sites/louvain-la-neuve), a walkable city, that offers a pleasant and dynamic living environment. The research project will be supervised by Patrick Watrin.

Required skills:

- A completed PhD in Computer Science, Machine Learning, NLP or a similar domain. - Excellent programming skills: - Python - TensorFlow/Keras or PyTorch - Linux (server administration) - Knowledge of the main supervised learning algorithms and deep learning algorithms is required - A good knowledge of the main NLP tools and algorithms is a plus - Strong research track record (publications, conferences, etc.) - Autonomy, teamwork, ability to understand and analyze needs, adaptability - Excellent command of the French language (at least C1) and good command of English (at least B2)

Conditions:

- Fixed-term contract of one year, renewable once - Salary based on experience, ranging from 4250€ to 4850€ (monthly, gross)

The position requires residency in Belgium. Candidates from outside the EU are responsible for obtaining the adequate visa and/or permits, with support from the UCLouvain.

How to apply:

- Deadline : March 31 - The application file should be sent electronically to Patrick Watrin ( patrick.watrin@uclouvain.be) and contain: - A detailed resume showing the adequate qualifications and skills, as well and the scientific/academic experiences and publications; - A cover letter in french, describing your interest for the role, how your profile complies with the project's needs, etc.; - A recommendation letter in french or in english.

The shortlisted candidates will be invited to participate in a remote videocall (details will be communicated in a timely manner).