Postdoctoral Researcher in Computational Sociology Applied to Health Misinformation (M/F)

Job Description

Missions

The PADMEN project (ANRS-MIE, 2026–2027) investigates the production, circulation, and reception of discourse on human metapneumovirus (HMPV) across online platforms (Facebook, Instagram, TikTok, Bluesky, and others). It draws on a multilingual (English, French), multi-platform corpus collected through platform APIs. The successful candidate will identify and build a typology of recurrent discourses, formalize them into a model, and then design and evaluate a detection model for misinformation-bearing discourse. The work combines corpus labeling, fine-tuning of language models, sociological interpretation informed by theories of belief diffusion, scientific dissemination, and strict compliance with the legal framework governing digital data (GDPR).
The objective is to identify the triggers of an infodemic: which discourses, which actors, and which events precede and fuel the information surge observed around a pathogen such as HMPV. The approach is iterative: the inductive identification of recurrent discourses informs the construction of the annotation scheme, which in turn provides the basis for corpus labeling and then for the fine-tuning and evaluation of detection models. It therefore combines a methodological dimension (state of the art, benchmarking, annotation, fine-tuning, and evaluation) with a theoretical one (sociology of belief diffusion and of disinformation).

Activity

• Conduct a literature review and a benchmark of existing methods for detecting and characterizing health misinformation, in order to establish the state of the art, the benchmark datasets used in the field, and the open scientific challenges.
• Prepare, clean, structure, and document the multilingual (English, French) HMPV corpus drawn from the various online platforms and from the scientometric databases used by the project.
• Identify the recurrent discourses running through this corpus, using semantic analysis methods (discourse analysis, graph-based methods, clustering, topic modeling, named entity recognition) combined with close qualitative reading.
• Build a typology of these discourses and situate them within their broader social and societal context.
• Design the annotation guidelines for the posts, organize and carry out the labeling of the English- and French-language datasets, and measure inter-annotator reliability.
• Train, fine-tune, and evaluate a model for detecting misinformation-bearing discourse, based on transformer architectures (BERT/RoBERTa, SBERT, generative language models) or on alternative approaches, and document its performance and limitations.
• Analyze the temporal and relational dynamics through which these discourses circulate (burst detection, network analysis) in order to identify the triggers of infodemic episodes.
• Carry out scientometric analyses (PubMed, OpenAlex, Dimensions) and compare the dynamics of scientific output on HMPV with those of the discourses circulating on online platforms.
• Interpret the results in light of sociological theories of belief diffusion, rumor, and disinformation.
• Write scientific articles in English and present the findings at conferences and symposia.
• Produce, version, and document the Python scripts, Jupyter notebooks, and accompanying documentation, so as to ensure the reproducibility of the analyses.
• Ensure strict compliance with the GDPR.
• Take part in the project's scientific meetings, in exchanges with the consortium partners, and in the associated scientific outreach activities.

Your Profil

Skills

You hold a PhD in sociology or computational social science, with experience in natural language processing (NLP) applied to the social sciences and, more specifically, to social media content, and you wish to place that experience at the service of a sociological approach to health-related questions.

Skills
• Knowledge of the core concepts of the sociology of belief diffusion, rumor, and disinformation.
• Scientometric skills: use of bibliographic databases (PubMed, OpenAlex, Web of Science), construction of indicators, and analysis of publication dynamics.
• Ability to conduct a literature review and a methodological benchmark identifying existing methods and open scientific challenges.
• Ability to build a typology of discourses that combines close analysis of the textual material with attention to the social and societal context: the actors involved, controversies, epidemic temporalities, and the reception of public health messaging.
• A proven track record in the analysis of large volumes of social media data during the COVID-19 pandemic.
• Command of Python programming and of the associated data analysis ecosystem (conda, pandas, scikit-learn, PyTorch, Hugging Face Transformers).
• Experience in labeling textual data: design of annotation schemes, measurement of intercoder agreement, and dataset quality management.
• Familiarity with transformer architectures, fine-tuning, and the evaluation of language models for classification tasks, in French as well as in English.
• Command of written scientific English and experience in writing and submitting articles to international peer-reviewed journals.
• Knowledge of the regulatory framework governing personal data: command of the GDPR and rigorous attention to compliance are essential.
• Ability to work in a multidisciplinary environment; autonomy, rigor, curiosity, and a collaborative outlook.
• Research experience in the health field (public health, epidemiology, infodemiology, health communication) would be an asset for the project.
• Familiarity with health-related conspiracy discourse would also be appreciated.

Your Work Environment

The successful candidate will be assigned to Team 2 of LAPSCO (Laboratoire de Psychologie Sociale et Cognitive – Laboratory of Social and Cognitive Psychology, UMR 6024, CNRS / Université Clermont Auvergne) in Clermont-Ferrand and will join the team of the PADMEN project, funded by ANRS Emerging Infectious Diseases. The postdoctoral researcher will work under the supervision of the principal investigator and will contribute to WP5, devoted to the modeling and detection of misinformation on human metapneumovirus (HMPV).

Constraints and risks

None

Compensation and benefits

Compensation

From €3,041 gross per month, commensurate with experience

Annual leave and RTT

44 jours

Remote Working practice and compensation

Pratique et indemnisation du TT

Transport

Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€

About the offer

About the CNRS

The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.

CNRS

The research professions