metadata

dataset_info:
  features:
    - name: review
      dtype: string
    - name: sentiment
      dtype: string
  splits:
    - name: train
      num_bytes: 75997554
      num_examples: 50000
  download_size: 48513136
  dataset_size: 75997554
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: mit
task_categories:
  - text-classification
language:
  - is
size_categories:
  - 1K<n<10K

IMDB movie reviews translated into Icelandic

Originally from https://github.com/cadia-lvl/sentiment-analysis.

Citations

@techreport{Johannsson2023,
  title = {Evaluating Icelandic Sentiment Analysis Models Trained on Translated Data},
  author = {Ólafur Aron Johannsson and Birkir Arndal and Eysteinn Örn},
  year = {2023},
  institution = {University of Reykjavik},
  department = {Department of Computer Science},
  month = {12},
  day = {15},
  note = {Supervised by Stefán Ólafsson and Hrafn Loftsson, Examined by Sigurjón Ingi Garðarsson}
}