Datasets:
Tasks:
Question Answering
Modalities:
Text
Formats:
json
Sub-tasks:
extractive-qa
Languages:
Catalan
Size:
1K - 10K
ArXiv:
License:
Update README.md
Browse files
README.md
CHANGED
@@ -55,11 +55,11 @@ task_ids:
|
|
55 |
|
56 |
### Dataset Summary
|
57 |
|
58 |
-
VilaQuAD, An extractive QA dataset for Catalan, from [VilaWeb](www.vilaweb.cat) newswire text.
|
59 |
|
60 |
This dataset contains 2095 of Catalan language news articles along with 1 to 5 questions referring to each fragment (or context).
|
61 |
|
62 |
-
VilaQuad articles are extracted from the daily [VilaWeb](www.vilaweb.cat) and used under [CC-by-nc-sa-nd](https://creativecommons.org/licenses/by-nc-nd/3.0/deed.ca) licence.
|
63 |
|
64 |
This dataset can be used to build extractive-QA and Language Models.
|
65 |
|
@@ -92,7 +92,7 @@ The dataset is in Catalan (`ca-CA`).
|
|
92 |
|
93 |
### Data Fields
|
94 |
|
95 |
-
Follows [Rajpurkar, Pranav et al., 2016](http://arxiv.org/abs/1606.05250) for SQuAD v1 datasets.
|
96 |
|
97 |
- `id` (str): Unique ID assigned to the question.
|
98 |
- `title` (str): Title of the VilaWeb article.
|
@@ -117,25 +117,25 @@ We created this dataset to contribute to the development of language models in C
|
|
117 |
|
118 |
### Source Data
|
119 |
|
120 |
-
- [VilaWeb site](www.vilaweb.cat)
|
121 |
|
122 |
#### Initial Data Collection and Normalization
|
123 |
|
124 |
The source data are scraped articles from archives of Catalan newspaper website [Vilaweb](https://www.vilaweb.cat).
|
125 |
|
126 |
-
From a the online edition of the newspaper [VilaWeb](https://www.vilaweb.cat), 2095 articles were randomnly selected. These headlines were also used to create a Textual Entailment dataset. For the extractive QA dataset, creation of between 1 and 5 questions for each news context was commissioned, following an adaptation of the guidelines from SQuAD 1.0 ([Rajpurkar, Pranav et al.
|
127 |
|
128 |
For compatibility with similar datasets in other languages, we followed as close as possible existing curation guidelines. We also created [another QA dataset with wikipedia](https://huggingface.co/datasets/projecte-aina/viquiquad) to ensure thematic and stylistic variety.
|
129 |
|
130 |
#### Who are the source language producers?
|
131 |
|
132 |
-
Professional journalists from the Catalan newspaper [VilaWeb](https://www.vilaweb.cat).
|
133 |
|
134 |
### Annotations
|
135 |
|
136 |
#### Annotation process
|
137 |
|
138 |
-
We comissioned the creation of 1 to 5 questions for each context, following an adaptation of the guidelines from SQuAD 1.0 ([Rajpurkar, Pranav et al.
|
139 |
|
140 |
#### Who are the annotators?
|
141 |
|
|
|
55 |
|
56 |
### Dataset Summary
|
57 |
|
58 |
+
VilaQuAD, An extractive QA dataset for Catalan, from [VilaWeb](https://www.vilaweb.cat/) newswire text.
|
59 |
|
60 |
This dataset contains 2095 of Catalan language news articles along with 1 to 5 questions referring to each fragment (or context).
|
61 |
|
62 |
+
VilaQuad articles are extracted from the daily [VilaWeb](https://www.vilaweb.cat/) and used under [CC-by-nc-sa-nd](https://creativecommons.org/licenses/by-nc-nd/3.0/deed.ca) licence.
|
63 |
|
64 |
This dataset can be used to build extractive-QA and Language Models.
|
65 |
|
|
|
92 |
|
93 |
### Data Fields
|
94 |
|
95 |
+
Follows [Rajpurkar, Pranav et al., (2016)](http://arxiv.org/abs/1606.05250) for SQuAD v1 datasets.
|
96 |
|
97 |
- `id` (str): Unique ID assigned to the question.
|
98 |
- `title` (str): Title of the VilaWeb article.
|
|
|
117 |
|
118 |
### Source Data
|
119 |
|
120 |
+
- [VilaWeb site](https://www.vilaweb.cat/)
|
121 |
|
122 |
#### Initial Data Collection and Normalization
|
123 |
|
124 |
The source data are scraped articles from archives of Catalan newspaper website [Vilaweb](https://www.vilaweb.cat).
|
125 |
|
126 |
+
From a the online edition of the newspaper [VilaWeb](https://www.vilaweb.cat), 2095 articles were randomnly selected. These headlines were also used to create a Textual Entailment dataset. For the extractive QA dataset, creation of between 1 and 5 questions for each news context was commissioned, following an adaptation of the guidelines from SQuAD 1.0 ([Rajpurkar, Pranav et al. (2016)](http://arxiv.org/abs/1606.05250)). In total, 6282 pairs of a question and an extracted fragment that contains the answer were created.
|
127 |
|
128 |
For compatibility with similar datasets in other languages, we followed as close as possible existing curation guidelines. We also created [another QA dataset with wikipedia](https://huggingface.co/datasets/projecte-aina/viquiquad) to ensure thematic and stylistic variety.
|
129 |
|
130 |
#### Who are the source language producers?
|
131 |
|
132 |
+
Professional journalists from the Catalan newspaper [VilaWeb](https://www.vilaweb.cat/).
|
133 |
|
134 |
### Annotations
|
135 |
|
136 |
#### Annotation process
|
137 |
|
138 |
+
We comissioned the creation of 1 to 5 questions for each context, following an adaptation of the guidelines from SQuAD 1.0 ([Rajpurkar, Pranav et al. (2016)](http://arxiv.org/abs/1606.05250)).
|
139 |
|
140 |
#### Who are the annotators?
|
141 |
|