|
--- |
|
task_categories: |
|
- text-generation |
|
language: |
|
- fo |
|
license: cc-by-4.0 |
|
--- |
|
|
|
# Dataset Card for Faroese_BLARK_small |
|
|
|
## Dataset Description |
|
|
|
All sentences are retrieved from: |
|
|
|
- **Paper:** |
|
|
|
Annika Simonsen, Sandra Saxov Lamhauge, Iben Nyholm Debess, and Peter Juel Henrichsen. 2022. Creating a Basic Language Resource Kit for Faroese. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 4637–4643, Marseille, France. European Language Resources Association. |
|
|
|
### Dataset Summary |
|
|
|
This dataset is a filtered version of the corpus (35.6 M tokens) first published as BLARK - Basic Language Resource Kit for Faroese. |
|
The pre-processing and filtering steps include: |
|
|
|
- Normalize format to utf-8 |
|
- Remove shorter sentences (less than 10 units, where units are separated by spaces) |
|
- Remove archaic Faroese |
|
- Remove separators ('\r', '\t', '\n') |
|
- Remove non standard formatting. Examples: '§§', ' | ', '**', ' • ', ' • ', '.- ', ': ?', '.?', '\xa0', '\xad', '_ _', '. .', etc. |
|
- Remove (most) numbered lists, of formats: 1), 1:, Stk. 1 etc. |
|
- Replace arbitrary number of question/exclamation marks and full-stops with 1. Example: !!!!!! -> ! |
|
- Remove websites that start with http |
|
- Remove sentences without (or with little) linguistic content. In practice: all sentences where more than half of the characters (excluding spaces) are number, punctuations and letters in caps-lock (acronyms and initials) |
|
- Remove duplicates |
|
|
|
### Supported Tasks and Leaderboards |
|
|
|
Suitable for MLM and CLM |
