GoldHamster Model
Model for text classification based on the GoldHamster corpus. Source code is available.
Model Description
Model pre-trained on PubMedBERT and fine-tuned on the GoldHamster corpus.
- Language(s) (NLP): English
- License: CC BY 3.0 DE
- Finetuned from model: https://huggingface.co/microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract
Model Sources
- Repository: https://github.com/mariananeves/goldhamster
- Paper: https://europepmc.org/article/ppr/ppr479254
Uses
Model for detecting our eight-label schema: invertebrates, in vivo, human, organs, primary cell lines, immortal cell lines, in silico, others. Predictions are on the document level.
Direct Use
[More Information Needed]
Results
Results are in terms of f-score.
| invertebrates | in vivo | human | organs | primary cell lines | immortal cell lines | in silico | others |
|---|---|---|---|---|---|---|---|
| 0.95 | 0.88 | 0.86 | 0.82 | 0.75 | 0.83 | 0.75 | 0.78 |
Citation
@misc {PPR:PPR479254,
Title = {Automatic classification of experimental models in biomedical literature to support searching for alternative methods to animal experiments},
Author = {Neves, Mariana and Klippert, Antonina and Knöspel, Fanny and Rudeck, Juliane and Stolz, Ailine and Ban, Zsofia and Becker, Markus and Diederich, Kai and Grune, Barbara and Kahnau, Pia and Ohnesorge, Nils and Pucher, Johannes and Schönfelder, Gilbert and Bert, Bettina and Butzke, Daniel},
DOI = {10.21203/rs.3.rs-1526055/v1},
Publisher = {Research Square},
Year = {2022},
URL = {https://doi.org/10.21203/rs.3.rs-1526055/v1},
}
Contact
Contact: https://mariananeves.github.io/