Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -7,21 +7,22 @@ language:
|
|
7 |
- ja
|
8 |
---
|
9 |
# Wikidata parallel descriptions en-ja
|
10 |
-
- Parallel corpus for machine translation generated from wikidata dump (2024-05-06).
|
11 |
- Currently we processed only English/Japanese pair.
|
12 |
-
- The `jsonl` file is ready-to-train by Hugging Face transformers translation
|
13 |
|
14 |
# Dataset Details
|
15 |
- https://www.wikidata.org/wiki/Wikidata:Database_download
|
16 |
|
17 |
## Dataset Creation
|
18 |
-
- As Wikidata description field
|
19 |
-
- We did word count based filtering by using Japanese tokenizer
|
20 |
- We also did deduplication.
|
21 |
|
22 |
## Bias, Risks, and Limitations
|
23 |
- The data is still noisy and might not be very parallel.
|
24 |
|
25 |
# License
|
26 |
-
- CC0
|
27 |
-
|
|
|
|
7 |
- ja
|
8 |
---
|
9 |
# Wikidata parallel descriptions en-ja
|
10 |
+
- Parallel corpus for machine translation generated from [wikidata dump](https://dumps.wikimedia.org/wikidatawiki/entities/) (2024-05-06).
|
11 |
- Currently we processed only English/Japanese pair.
|
12 |
+
- The `jsonl` file is ready-to-train by Hugging Face transformers trainer for translation tasks.
|
13 |
|
14 |
# Dataset Details
|
15 |
- https://www.wikidata.org/wiki/Wikidata:Database_download
|
16 |
|
17 |
## Dataset Creation
|
18 |
+
- As Wikidata description field does not represent exact direct translation, filtering is required for machine translation task.
|
19 |
+
- We did word count based filtering by using Japanese tokenizer MeCab and did some word-based filtering like `ウィキメディア`
|
20 |
- We also did deduplication.
|
21 |
|
22 |
## Bias, Risks, and Limitations
|
23 |
- The data is still noisy and might not be very parallel.
|
24 |
|
25 |
# License
|
26 |
+
- [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/)
|
27 |
+
- Note : All structured data in Wikidata is also licensed under CC0.
|
28 |
+
- https://www.wikidata.org/wiki/Wikidata:Licensing
|