Update README.md
Browse files
README.md
CHANGED
|
@@ -45,6 +45,9 @@ This dataset has generated from:
|
|
| 45 |
- "Oraciones en Español traducidas a Quechua" by "Tatoeba" - [available here](https://tatoeba.org/es/sentences/search?from=spa&query=&to=que)
|
| 46 |
- "AmericasNLP 2021 Shared Task on Open Machine Translation" by "americasnlp2021" - [available here](https://github.com/AmericasNLP/americasnlp2021/tree/main/data/quechua-spanish/parallel_data/es-quy)
|
| 47 |
|
|
|
|
|
|
|
|
|
|
| 48 |
## Considerations for Using the Data
|
| 49 |
|
| 50 |
This is a first version of the dataset, we expected improve it over time and especially to neutralize the biblical themes.
|
|
|
|
| 45 |
- "Oraciones en Español traducidas a Quechua" by "Tatoeba" - [available here](https://tatoeba.org/es/sentences/search?from=spa&query=&to=que)
|
| 46 |
- "AmericasNLP 2021 Shared Task on Open Machine Translation" by "americasnlp2021" - [available here](https://github.com/AmericasNLP/americasnlp2021/tree/main/data/quechua-spanish/parallel_data/es-quy)
|
| 47 |
|
| 48 |
+
### Data cleaning
|
| 49 |
+
- The dataset was manually cleaned during compilation, as some words of one language were related to several words of the other language.
|
| 50 |
+
|
| 51 |
## Considerations for Using the Data
|
| 52 |
|
| 53 |
This is a first version of the dataset, we expected improve it over time and especially to neutralize the biblical themes.
|