cjvt/sloTS
To increase the accessibility and diversity of easy reading in Slovenian and to create a prototype system that automatically simplifies texts in Slovenian, we prepared a dataset for the Slovenian language that contains aligned simple and complex sentences, which can be used for further development of models for simplifying texts in Slovenian. Dataset is a .json file that usually contains one complex ("kompleksni") and one simplified sentence ("enostavni") per row. However, if a complex sentence contains a lot of information we translated this sentence into more than one simplified sentences. Vice versa, more complex sentences can be translated into one simplified sentence if some information is given through more than one complex sentences but we summarised them into one simplified one.
Dataset Card for SloTS
Dataset Summary
SloTS is a sentence simplification dataset containing 973 pairs of complex and simplified sentences. In some cases one complex sentence is translated into multiple simplified sentences, or more complex sentences are translated into one simplified sentence.
Languages
Slovenian.
Dataset Structure
Data Instances
A sample instance from the dataset:
{
'complex': 'Vsa vas je dobro vedela, da ga na svetu ni hudobnejšega človeka od Vrbarjevega Matevža .',
'simple': 'Matevž je bil zelo hudoben človek .'
}Data Fields
- 'complex': sentence in its complex form;
- 'simple': sentence in its simplified form.
Additional Information
Dataset Curators
Gorenc, Sabina and Robnik-Šikonja, Marko
Licensing Information
CC BY 4.0
Citation Information
@misc{sloTS,
title = {Slovene text simplification dataset {SloTS}},
author = {Gorenc, Sabina and Robnik-{\v S}ikonja, Marko},
url = {http://hdl.handle.net/11356/1682},
note = {Slovenian language resource repository {CLARIN}.{SI}},
copyright = {Creative Commons - Attribution 4.0 International ({CC} {BY} 4.0)},
year = {2022}
}Contributions
Thanks to Hana Skitek for adding this dataset.