README.md

---
language: bn
license: cc-by-nc-sa-4.0
extra_gated_prompt: 'By accessing this dataset, you agree to abide by the BigScience
  Ethical Charter. The charter can be found at:

  https://hf.co/spaces/bigscience/ethical-charter'
extra_gated_fields:
  I have read and agree to abide by the BigScience Ethical Charter: checkbox
---
ROOTS Subset: roots_indic-bn_bangla_sentiment_classification_datasets

# Bangla Sentiment Classification Datasets

- Dataset uid: `bangla_sentiment_classification_datasets`

### Description

Multiple sentiment classification datasets for Bengali, which can also be used for training LMs. The Datasets are the following:

ABSA_datasets -- This dataset has developed to per­form aspect ­based sentiment analysis task in Bangla.
License: CC BY 4.0

SAIL_data -- This dataset, consists of tweet posts, has developed in the Shared task on Sentiment Analysis in Indian Languages (SAIL) 2015. [2]
License: Non-profit

multichannel_bsentiment -- This data was collected from several newspapers, TV news, books, blogs, and social me­dia.
License: NA (N.B.: We have the permission to share this dataset.)

youtube_sentiment -- This dataset was devel­oped by extracting comments from various YouTube videos.
License: NA

### Homepage

https://github.com/banglanlp/bnlp-resources/tree/main/sentiment

### Licensing

- research use
- cc-by-4.0: Creative Commons Attribution 4.0 International

The work is licensed under https://creativecommons.org/licenses/by-nc/4.0/. However, for each dataset, please see the license information associated with it. Any private dataset can be accessed by contacting with the respective authors.


### Speaker Locations

- Southern Asia
- Bangladesh


### Sizes

- 0.0002 % of total
- 0.0074 % of indic-bn

### BigScience processing steps

#### Filters applied to: indic-bn

- dedup_document
- dedup_template_soft
- filter_remove_empty_docs
- filter_small_docs_bytes_300