README.md

metadata

task_categories:
  - text-classification
task_ids:
  - topic-classification
language:
  - de
pretty_name: Scientific Subset of 10kGNAD
source_datasets:
  - community-datasets/gnad10
multilinguality:
  - monolingual
size_categories:
  - n<1K
dataset_info:
  - config_name: default
    features:
      - name: label
        dtype:
          class_label:
            names:
              '0': Web
              '1': Panorama
              '2': International
              '3': Wirtschaft
              '4': Sport
              '5': Inland
              '6': Etat
              '7': Wissenschaft
              '8': Kultur
      - name: text
        dtype: string
  - config_name: binary
    features:
      - name: label
        dtype:
          class_label:
            names:
              '0': Wissenschaft
              '1': Nicht-Wissenschaft
      - name: text
        dtype: string
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train.csv
      - split: validation
        path: data/validation.csv
      - split: test
        path: data/test.csv
    sep: ;
    quotechar: ''''
  - config_name: binary
    data_files:
      - split: train
        path: binary/train.csv
      - split: validation
        path: binary/validation.csv
      - split: test
        path: binary/test.csv
    sep: ;
    quotechar: ''''