EUVIMEDSkopri EUVIMED
Open access · CC BY 4.0

Using large language models to extract plant functional traits from unstructured text

Viktor Domazetoski iD, Holger Kreft iD, Helena Bestova, Philipp Wieder, Radoslav Koynov, Alireza Zarei, Patrick Weigelt iD

Applications in Plant Sciences · 2025-05 · 13 (3)

DOI: 10.1002/aps3.70011 ↗

At a glance

Journal article in Applications in Plant Sciences, published online on 2025-06-03. Written by Viktor Domazetoski and 6 co-authors. Institution: Department of Biodiversity, Macroecology, and Biogeography University of Göttingen Göttingen Germany. Cited 16 times so far (Crossref); 69 references. The full text is freely available (open access, CC BY 4.0).

AstrattSubjectsDettalji biblijografiċiCiteRelated studies

Abstract (original language)

Premise Functional plant ecology seeks to understand how functional traits govern species distributions, community assembly, and ecosystem functions. While global trait datasets have advanced the field, substantial gaps remain, and extracting trait information from text in books, research articles, and online sources via machine learning offers a valuable complement to costly field campaigns.

Methods

We propose a natural language processing pipeline that extracts traits from unstructured species descriptions by using classification models for categorical traits and question‐answering models for numerical traits. The pipeline's performance is evaluated on two large databases with over 50,000 species descriptions, utilizing approaches ranging from a keyword search to large language models.

Results

Our final optimized pipeline used a transformer architecture and obtained a mean precision of 90.8% (range 81.6–97%) and a mean recall of 88.6% (77.4–97%) across five categorical traits, representing a 9.83% increase in precision and 42.35% increase in recall over a regular expression‐based approach. The question‐answering model yielded a normalized mean absolute error of 10.3% averaged across three numerical traits.

Discussion

The natural language processing pipeline we propose has the potential to facilitate the digitization and extraction of large amounts of plant functional trait information residing in scattered textual descriptions.

Subjects

Author keywords

automatic trait extractionbiodiversityfunctional plant ecologylarge language modelsnatural language processingvascular plants

Authors and institutions

  • Department of Biodiversity, Macroecology, and Biogeography University of Göttingen Göttingen Germany.
  • Campus-Institute Data Science Göttingen Germany.
  • Centre of Biodiversity and Sustainable Land Use University of Göttingen Göttingen Germany.
  • Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen Göttingen Germany.
  • Department of Environmental Science, Radboud Institute for Biological and Environmental Sciences Radboud University Heyendaalseweg 135 Nijmegen 6525AJ The Netherlands.

ORCID profiles: Viktor Domazetoski, Holger Kreft, Patrick Weigelt

Cite

APA

Domazetoski, V., Kreft, H., Bestova, H., Wieder, P., Koynov, R., Zarei, A., & Weigelt, P. (2025). Using large language models to extract plant functional traits from unstructured text. Applications in Plant Sciences, 13(3). https://doi.org/10.1002/aps3.70011

Vancouver

Domazetoski V, Kreft H, Bestova H, Wieder P, Koynov R, Zarei A, et al. Using large language models to extract plant functional traits from unstructured text. Applications in Plant Sciences. 2025;13(3). doi:10.1002/aps3.70011

BibTeX
@article{Domazetoski2025,
  author = {Domazetoski, Viktor and Kreft, Holger and Bestova, Helena and Wieder, Philipp and Koynov, Radoslav and Zarei, Alireza and Weigelt, Patrick},
  title = {Using large language models to extract plant functional traits from unstructured text},
  journal = {Applications in Plant Sciences},
  year = {2025},
  volume = {13},
  number = {3},
  doi = {10.1002/aps3.70011}
}
Start for free →