Datasets:
Languages:
Indonesian
Tags:
question-answering
holylovenia
commited on
Commit
•
c6a0966
1
Parent(s):
76d91b7
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -3,4 +3,70 @@ tags:
|
|
3 |
- question-answering
|
4 |
language:
|
5 |
- ind
|
6 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
- question-answering
|
4 |
language:
|
5 |
- ind
|
6 |
+
---
|
7 |
+
|
8 |
+
TyDiQA dataset is collected from Wikipedia articles with human-annotated question and answer pairs covering 11 languages.
|
9 |
+
The question-answer pairs are collected for each language without using translation services.
|
10 |
+
IndoNLG uses the Indonesian data from the secondary Gold passage task of the original TyDiQA dataset and
|
11 |
+
randomly split off 15% of the training data and use it as the test set.
|
12 |
+
|
13 |
+
|
14 |
+
## Dataset Usage
|
15 |
+
|
16 |
+
Run `pip install nusacrowd` before loading the dataset through HuggingFace's `load_dataset`.
|
17 |
+
|
18 |
+
## Citation
|
19 |
+
|
20 |
+
```@article{clark-etal-2020-tydi,
|
21 |
+
title = "{T}y{D}i {QA}: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages",
|
22 |
+
author = "Clark, Jonathan H. and
|
23 |
+
Choi, Eunsol and
|
24 |
+
Collins, Michael and
|
25 |
+
Garrette, Dan and
|
26 |
+
Kwiatkowski, Tom and
|
27 |
+
Nikolaev, Vitaly and
|
28 |
+
Palomaki, Jennimaria",
|
29 |
+
journal = "Transactions of the Association for Computational Linguistics",
|
30 |
+
volume = "8",
|
31 |
+
year = "2020",
|
32 |
+
address = "Cambridge, MA",
|
33 |
+
publisher = "MIT Press",
|
34 |
+
url = "https://aclanthology.org/2020.tacl-1.30",
|
35 |
+
doi = "10.1162/tacl_a_00317",
|
36 |
+
pages = "454--470",
|
37 |
+
}
|
38 |
+
|
39 |
+
@inproceedings{cahyawijaya-etal-2021-indonlg,
|
40 |
+
title = "{I}ndo{NLG}: Benchmark and Resources for Evaluating {I}ndonesian Natural Language Generation",
|
41 |
+
author = "Cahyawijaya, Samuel and
|
42 |
+
Winata, Genta Indra and
|
43 |
+
Wilie, Bryan and
|
44 |
+
Vincentio, Karissa and
|
45 |
+
Li, Xiaohong and
|
46 |
+
Kuncoro, Adhiguna and
|
47 |
+
Ruder, Sebastian and
|
48 |
+
Lim, Zhi Yuan and
|
49 |
+
Bahar, Syafri and
|
50 |
+
Khodra, Masayu and
|
51 |
+
Purwarianti, Ayu and
|
52 |
+
Fung, Pascale",
|
53 |
+
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
|
54 |
+
month = nov,
|
55 |
+
year = "2021",
|
56 |
+
address = "Online and Punta Cana, Dominican Republic",
|
57 |
+
publisher = "Association for Computational Linguistics",
|
58 |
+
url = "https://aclanthology.org/2021.emnlp-main.699",
|
59 |
+
doi = "10.18653/v1/2021.emnlp-main.699",
|
60 |
+
pages = "8875--8898"
|
61 |
+
}
|
62 |
+
```
|
63 |
+
|
64 |
+
## License
|
65 |
+
|
66 |
+
Creative Common Attribution Share-Alike 4.0 International
|
67 |
+
|
68 |
+
## Homepage
|
69 |
+
|
70 |
+
### NusaCatalogue
|
71 |
+
|
72 |
+
For easy indexing and metadata: [https://indonlp.github.io/nusa-catalogue](https://indonlp.github.io/nusa-catalogue)
|