Observed arrival · 2026-09-15
Clément Longeac’s French Oncology NLP Lab Notebook
A research portfolio documenting hybrid NLP systems for extracting biomarkers, staging information, and treatment details from French oncology reports.
Field notes
The portfolio describes DEMNE/DuraXeLL as a staged clinical-NLP pipeline rather than a single-model system: rules handle straightforward entities, a CRF resolves sequence context, and DrBERT-7GB handles domain-heavy cases before an LLM arbitration pass. Its stated scope is French oncology text, with three corpora and seven priority biomarker categories. The author presents the work alongside an ESIEE Paris engineering path and plans for biomedical-AI doctoral research.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue