text-classification easy-language plain-language leichte-sprache einfache-sprache text-complexity

Model Card for text-complexity-classification

The model classifies texts into the language complexity classes (German language):

The underlying corpus was trained on the basis of over 300,000 texts of the mentioned language categories. Freely available websites served as sources. Thematic diversity was taken into account when selecting the sources.

Evaluation

How to use

>>> from transformers import pipeline
>>> classifier = pipeline(model="krupper/text-complexity-classification")
>>> classifier("Bei Kleinkindern unter 2 Jahren liegen nur begrenzte Erfahrungen zur Pharmakokinetik vor.")
[{'label': 'special_language', 'score': 0.999923825263977}]

DOI: https://doi.org/10.57967/hf/0131