Postdoctoral position: Scientific corpus analysis and semantic
annotation

Location: Marie and Louis Pasteur University, CRIT Laboratory, Besançon, France

Duration: 24 months

Preferred starting date: between June and October 2026 (to be specified)

Context

As part of an application to the ATRACT programme (Accueil de Talents
de la Recherche et Appui à la Coopération sur le Territoire), the CRIT
laboratory is seeking to recruit a high-level early-career researcher
for a 24-month postdoctoral position in Natural Language
Processing. The objective is to strengthen the laboratory's work on
large-scale scientific discourse analysis, in continuity with projects
on the modelling of scientific uncertainty across disciplines (ANR
InSciM project) and, more generally, on annotation and information
extraction from scientific corpora. The research will be conducted
under the supervision of Prof. Iana Atanassova.

Scientific focus

The successful candidate will work on large-scale scientific corpora,
with a focus on:

- Analysis of scientific corpora (articles, reports, preprints, etc.)
using NLP methods, including linguistically informed, machine learning
and deep learning approaches.

- Analysis of scientific discourse across disciplines, with
  applications to epistemology.

- Development or adaptation of methods for text mining, information
extraction and/or semantic modelling (e.g. contextual embeddings,
language models, classification methods).

- Design of reusable resources and tools (annotation schemes,
annotated corpora, models, scripts) in connection with ongoing
projects in the laboratory.

The project may be adapted to the profile and research interests of
the successful candidate, within the scope of the laboratory's
thematic priorities.

Profile

The candidate should:

- Hold a PhD in Natural Language Processing, Computational
Linguistics, Computer Science, Data Science, or a closely related
field.

- Have defended their PhD less than 10 years ago.

- Have a strong scientific track record (international publications,
conference presentations, participation in research projects).

- Demonstrate expertise in several of the following areas:

- NLP applied to scientific or specialised texts.

- Manual and/or semi-automatic corpus annotation (annotation schemes,
guidelines, coordination of annotators).

- Rule-based and linguistically informed methods and algorithms.

- Neural models, embeddings, LLMs, etc.

- Information extraction and discourse analysis.

- Have solid scientific programming skills (Python, NLP/ML ecosystem
such as spaCy, HuggingFace, scikit-learn, etc.).

- Currently be affiliated with an institution located outside the
Bourgogne-Franche-Comté region and be willing to join UMLP (Besançon)
for the duration of the contract.

- Have an excellent command of scientific English and French. Both
languages will be used in project meetings.

Experience in epistemology will be particularly appreciated.

The position is part of an application currently being prepared for
the ATRACT programme of the Bourgogne-Franche-Comté Region. The
successful candidate will benefit from a multidisciplinary scientific
environment within the CRIT laboratory and from regional, national and
international collaborations.

Application

Interested candidates are invited to send the following documents by e-mail before 28 December 2025:

- A detailed CV (including a full list of publications).
- A cover letter describing:
- Main research contributions.
- Possible connections with the research topics described above and
with the proposed project.
- The contact details of two referees (name, affiliation, e-mail) who
may be contacted for reference letters during the week of 5-9 January.

Applications should be sent to: iana.atanassova@umlp.fr

Suggested e-mail subject line: Application ATRACT Postdoctoral
Position - [Last name First name]

Applications will be reviewed on a rolling basis. Online interviews
may be organised in early January, in view of submitting the ATRACT
proposal before 10 January.