metadata

dataset_info:
  features:
    - name: text
      dtype: string
    - name: label
      dtype: string
  splits:
    - name: train
      num_bytes: 2312958
      num_examples: 11807
  download_size: 918814
  dataset_size: 2312958
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - text-classification
language:
  - bn

Bengali Sentiment Analysis

Context

The dataset contains 3307 Negative reviews and 8500 Positive reviews collected and manually annotated from Youtube Bengali drama.

Acknowledgements

Sazzed, Salim (2021), “Bangla ( Bengali ) sentiment analysis classification benchmark dataset corpus”, Mendeley Data, V4, doi: 10.17632/p6zc7krs37.4