metadata

license: cc-by-nc-sa-3.0
task_categories:
  - question-answering
language:
  - ar
  - en
  - es
  - fr
  - ko
  - ru
  - vi
  - zh
size_categories:
  - n<1K
configs:
  - config_name: english
    default: true
    data_files:
      - split: test
        path: data/english.csv
  - config_name: arabic
    data_files:
      - split: test
        path: data/arabic.csv
  - config_name: chinese
    data_files:
      - split: test
        path: data/chinese.csv
  - config_name: french
    data_files:
      - split: test
        path: data/french.csv
  - config_name: korean
    data_files:
      - split: test
        path: data/korean.csv
  - config_name: korean
    data_files:
      - split: test
        path: data/korean.csv
  - config_name: russian
    data_files:
      - split: test
        path: data/russian.csv
  - config_name: spanish
    data_files:
      - split: test
        path: data/spanish.csv
  - config_name: vietnamese
    data_files:
      - split: test
        path: data/vietnamese.csv

Usage

import datasets

langs = ['arabic', 'chinese', 'english', 'french', 'korean', 'russian', 'spanish', 'vietnamese']
data = datasets.load_dataset('xhluca/publichealth-qa', split='test', name=langs[0])

About

This dataset contains question and answer pairs sourced from Q&A pages and FAQs from CDC and WHO pertaining to COVID-19. They were produced and collected between 2019-12 and 2020-04. They were originally published as an aggregated Kaggle dataset.

License

CDC data is licensed under CC-BY 3.0 and WHO is licensed under cc-by-nc-sa-3.0.

Source

This data was originally included in the COVID-QA dataset, where it was known as the multilingual split. The files in this updated repository were generated using the publichealth-qa repository.