Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -44,17 +44,17 @@ language:
|
|
44 |
|
45 |
ソース: [yahoojapan/JGLUE on GitHub](https://github.com/yahoojapan/JGLUE/tree/main)
|
46 |
|
47 |
-
|
48 |
|
49 |
> JCommonsenseQA is a Japanese version of CommonsenseQA (Talmor+, 2019), which is a multiple-choice question answering dataset that requires commonsense reasoning ability.
|
50 |
> It is built using crowdsourcing with seeds extracted from the knowledge base ConceptNet.
|
51 |
|
52 |
|
53 |
-
|
54 |
|
55 |
[Creative Commons Attribution Share Alike 4.0 International](https://github.com/yahoojapan/JGLUE/blob/main/LICENSE)
|
56 |
|
57 |
-
|
58 |
|
59 |
```
|
60 |
@article{栗原 健太郎2023,
|
@@ -95,13 +95,13 @@ language:
|
|
95 |
```
|
96 |
|
97 |
|
98 |
-
|
99 |
|
100 |
-
|
101 |
|
102 |
-
|
103 |
|
104 |
-
- `q_id` (`str`):
|
105 |
-
- `question` (`str`):
|
106 |
-
- `choice{0..4}` (`list[str]`):
|
107 |
-
- `label` (`int`):
|
|
|
44 |
|
45 |
ソース: [yahoojapan/JGLUE on GitHub](https://github.com/yahoojapan/JGLUE/tree/main)
|
46 |
|
47 |
+
# JCommonsenseQA
|
48 |
|
49 |
> JCommonsenseQA is a Japanese version of CommonsenseQA (Talmor+, 2019), which is a multiple-choice question answering dataset that requires commonsense reasoning ability.
|
50 |
> It is built using crowdsourcing with seeds extracted from the knowledge base ConceptNet.
|
51 |
|
52 |
|
53 |
+
## Licensing Information
|
54 |
|
55 |
[Creative Commons Attribution Share Alike 4.0 International](https://github.com/yahoojapan/JGLUE/blob/main/LICENSE)
|
56 |
|
57 |
+
## Citation Information
|
58 |
|
59 |
```
|
60 |
@article{栗原 健太郎2023,
|
|
|
95 |
```
|
96 |
|
97 |
|
98 |
+
# Subsets
|
99 |
|
100 |
+
## default
|
101 |
|
102 |
+
[datasets/jcommonsenseqa-v1.1](https://github.com/yahoojapan/JGLUE/tree/main/datasets/jcommonsenseqa-v1.1)
|
103 |
|
104 |
+
- `q_id` (`str`): 質問を一意識別するための ID
|
105 |
+
- `question` (`str`): 質問文
|
106 |
+
- `choice{0..4}` (`list[str]`): 選択肢(5つ)
|
107 |
+
- `label` (`int`): `choices` に対応した正解選択肢のインデックス
|