UnScientify - Scientific Uncertainty Extractor

Open-source tool for the detection and categorisation of scientific uncertainty

UnScientify - Scientific Uncertainty Extractor

UnScientify is a linguistic knowledge-based tool for the automatic annotation of scientific articles and for the detection and categorisation of scientific uncertainty. It relies on spaCy and on a set of lexical and morphosyntactic patterns, and was developed as part of the ANR InSciM project (ANR-21-CE38-0003-01).

Code: github.com/Project-InSciM/Scientific-Uncertainty-Extractor (BSD 3-Clause licence)

Contributors: Panggih Kusuma Ningrum, Nicolas Gutehrlé, Iana Atanassova.

Reference: Ningrum, P. K., Mayr, P., Smirnova, N., & Atanassova, I. (2025). Annotating scientific uncertainty: A comprehensive model using linguistic patterns and comparison with existing approaches. Journal of Informetrics, 19(2), 101661. doi:10.1016/j.joi.2025.101661