Datasets:

Modalities:
Text
Formats:
parquet
Languages:
Slovak
ArXiv:
Libraries:
Datasets
pandas
License:
drchajan commited on
Commit
e3c2e2f
1 Parent(s): 35a71a1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +22 -0
README.md CHANGED
@@ -29,4 +29,26 @@ configs:
29
  path: data/train-*
30
  - split: validation
31
  path: data/validation-*
 
 
 
 
 
 
 
 
32
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
29
  path: data/train-*
30
  - split: validation
31
  path: data/validation-*
32
+ license: mit
33
+ task_categories:
34
+ - text2text-generation
35
+ language:
36
+ - sk
37
+ pretty_name: QA2D-sk
38
+ size_categories:
39
+ - 10K<n<100K
40
  ---
41
+
42
+ Slovak version of the Question to Declarative Sentence ([QA2D](https://huggingface.co/datasets/domenicrosati/QA2D)). Machine-translated using [DeepL](https://www.deepl.com]) service.
43
+
44
+ For more information, see our [Pipeline and Dataset Generation for Automated Fact-checking in Almost Any Language](https://arxiv.org/abs/2312.10171) paper.
45
+ Currently in review for [NCAA](https://link.springer.com/journal/521) journal.
46
+
47
+ ```bibtex
48
+ @article{drchal2023pipeline,
49
+ title={Pipeline and Dataset Generation for Automated Fact-checking in Almost Any Language},
50
+ author={Drchal, Jan and Ullrich, Herbert and Mlyn{\'a}{\v{r}}, Tom{\'a}{\v{s}} and Moravec, V{\'a}clav},
51
+ journal={arXiv preprint arXiv:2312.10171},
52
+ year={2023}
53
+ }
54
+ ```