Datasets:

Modalities:
Text
Formats:
json
Languages:
Romansh
Libraries:
Datasets
pandas
License:
quotidiana / README.md
jvamvas's picture
Splitting the single config with two splits into two configs, each with one split.
57b457d
metadata
license: cc-by-4.0
language:
  - rm
configs:
  - config_name: '1997_2008'
    data_files:
      - split: train
        path: 1997-2008.jsonl
  - config_name: '2021_2025'
    data_files:
      - split: train
        path: 2021-2025.jsonl

La Quotidiana news articles in Romansh

This dataset has two subsets:

  • 1997_2008: Articles published by La Quotidiana between 1997 and 2008. The articles have been retrieved from https://github.com/ProSvizraRumantscha/corpora and extracted from the original XML format.
  • 2021_2025: Articles published by La Quotidiana between 2021 and 2025. These articles have been extracted from WXR WordPress export files.

Annotation of Romansh Varieties

The dataset includes articles in different Romansh varieties: Rumantsch Grischun (rm-rumgr), Sursilvan (rm-sursilv), Sutsilvan (rm-sutsilv), Surmiran (rm-surmiran), Puter (rm-puter), and Vallader (rm-vallader).

  • For the 1997_2008 subset, the variety has been automatically annotated using an SVM classifier trained on the 2021_2025 config, among other data. More information here.
  • For the 2021_2025 subset, the variety is based on manual processing by La Quotidiana editors.

License

CC BY 4.0 License.
© La Quotidiana.