UCLouvain is looking for a post-doctoral researcher in natural language processing or computer science
- Full-time (100%) post-doctoral funding for a period of approximately 18 months
- for the Centre de traitement automatique du langage (CENTAL), of the Institut langage et communication (ILC) at UCLouvain (Louvain-la-Neuve)
- start date: 1 October 2024 (negotiable)
This post-doctoral research contract is part of the European project iRead4Skills - Intelligent reading improvement system for the development of fundamental and transversal skills (https://iread4skills.com/fr/). iREad4Skills is a research and innovation project funded by the European Commission: HORIZON-CL2-2022-TRANSFORMATIONS-01-07, carried out in collaboration with 8 partners in Spain, Portugal, and Belgium. The researcher recruited will be part of the team at the Centre de traitement automatique du langage (CENTAL, https://uclouvain.be/fr/instituts-recherche/ilc/cental) at UCLouvain in Louvain-la-Neuve (Belgium).
The iRead4Skills project aims to promote the development of reading skills among adults who are poor readers through an innovative intelligent system that assesses the complexity of texts and suggests reading adapted to the user's level. The system can also be used by trainers to select or adapt texts to an appropriate level of complexity for their students. Finally, by producing a precise diagnosis of where texts are difficult, the system could be used to help simplifying texts. The project therefore lies at the intersection between the fields of legibility, automatic text simplification and teaching reading to poor readers. A timetable of the project's objectives is available on the website (https://iread4skills.com/fr/activites/#calendrier-projet).
This project has a number of advantages. Firstly, it is rooted in a laboratory whose research into the automatic assessment of text readability and the automatic simplification of texts is internationally recognised. Secondly, it is part of a solid consortium bringing together experts representing the different disciplines needed to successfully complete the project.
Within the project, the role of the postdoctoral researcher hired will be to be responsible for the part of the research that concerns the development of tools for a fine annotation of the difficulties in texts, i.e. (1) to develop one or more machine learning model algorithms (classical and deep learning) capable of identifying the parts of a text that are difficult to read (complex words, difficult syntactic structures, etc.). To do this, it will be able to rely on a corpus of texts whose overall difficulty and difficult textual elements will have been annotated by teachers and weak readers. As the amount of data is relatively limited, the person involved will also be led to explore avenues based on large language models and few-shot setting. Secondly, (2) on the basis of this annotation, the researcher will participate in the development of a web tool to identify difficult-to-read segments in a text. The researcher will also be responsible for disseminating the results obtained via scientific publications and actively collaborating with a doctoral student who is also involved in the project and in charge of developing readability formulas. A number of missions to partner universities are planned, in order to ensure good collaboration within the consortium. A budget is available for conference travel for scientific publications related to the project.
Working environment
CENTAL is part of the Institut Langage & Communication (https://uclouvain.be/fr/instituts-recherche/ilc), which is part of UCLouvain. This university is located in Louvain-la-Neuve (https://uclouvain.be/fr/sites/louvain-la-neuve), a pedestrianised, pleasant and very dynamic city. The research project will be carried out under the supervision of Prof. Thomas François (https://cental.uclouvain.be/team/tfrancois/), an expert in automatic readability assessment and automatic text simplification.
Qualifications and skills
The candidate should have the following qualifications:
- Hold a PhD in Computational Linguistics, Natural Language Processing (NLP), or Computer Science (option in Artificial Intelligence).
- In view of the tasks to be carried out, excellent computer skills are required:
- programming languages: Python (mastery)
- scikit-learn, pandas, TensorFlow or PyTorch.
- web technologies: HTML, Javascript, CSS, Flask, MySQL and PostgreSQL
- systems: Linux (server management)
- Experience in creating modules in python released on github (or equivalent) is a plus.
- Knowledge of the main supervised learning algorithms and of deep neural networks is required.
- Excellent command of French (B2 level minimum and C1 preferred) and good knowledge of English (B2 level minimum).
- Excellent research and academic profile (publications, conferences, etc.)
- Good knowledge of the main NLP tools and algorithms, in particular for French.
- Research experience in one of the following areas would be a plus: computer-assisted language learning, readability, automatic essay scoring, language assessment, etc.
- Autonomy, ability to work in a team, ability to listen and analyse needs, responsiveness.
Recruitment conditions
Depending on the profile of the successful candidate, the commitment may take the form of a post-doctoral fellowship of approximately 18 months or a post-doctoral research contract of approximately 18 months (with an end date of 28 February 2026).
- The net amount of the fellowship or research contract starts at €3,000 per month and may be increased according to seniority.
- The researcher must be domiciled in Belgium for legal reasons.
Application file
Application deadline: 1 September 2024
If you are interested in this position, please send your application to Thomas François by email (thomas.francois@uclouvain.be). This should include :
1. a detailed curriculum vitae in French or English, including the various qualifications and skills required, details of your academic career (years of study, degrees obtained, honours), as well as scientific publications and other academic and scientific achievements;
2. a covering letter in French, describing your interest in the position ; how your profile meets the description of the position and the objectives of the project, etc. (maximum 2 pages);
3. one or two letters of reference in French or English, in particular from your thesis supervisor.
Successful candidates will be invited to take part in a video-conference interview, the details of which will be sent to them by email.
Further information :
Questions about the post or the application procedure should be sent by e-mail to Thomas François (thomas.francois@uclouvain.be) before 1 September.