language:

This T5 model is designed to correct the word order inside sentence sections. Sentences are split into sections based on commas and conjunctions.

Imagine we have the following Slovenian text:

The model might return the following text (note: predictions chosen for demonstration/explanation, not reproducibility!):

We observe that in the input sentence, the sentence sections Popravi model besedilo and v katerem vrstni je red nekaterih besed napačen are written with incorrect word order, so our model corrects word order of words Popravi model and je vrstni.

Testing the model with generated test sets provides the following result (combining detection and correction of words with incorrect word order):

The authors acknowledge the financial support from the Slovenian Research and Innovation Agency - research core funding No. P6-0411: Language Resources and Technologies for Slovene and research project No. J7-3159: Empirical foundations for digitally-supported development of writing skills.

Thanks to Martin Božič, Marko Robnik-Šikonja and Špela Arhar Holdt for developing these models.