/swahili-sentiment-analysis-dataset

Dataset for training Sentiment analysis models

MIT LicenseMIT

sentiment-analysis-dataset [Swahili]

Swahili sentiment analysis dataset is the dataset training binary (positive or negative) sentiment analysis model.

This data you see in swahili.csv is but results of cleaning and back-translation of data found on an opensource repository titled Swahili-sentiment-analysis by jinamizi.

This is the same data that was used to train Spark NLP | Swahili Sentiment analysis model.

Here a preview of the data

>>> import pandas as pd
>>> dataset = pd.read_csv('swahili.csv', usecols=['text', 'labels'])
>>> dataset.head()
                                 text    labels
0  team 2019merimera alikuwa takataka  negative
1                      sijafurahishwa  negative
2                       kubuni dosari  negative
3   bila kusema nilipoteza pesa zangu  negative
4        sema kupoteza pesa na wakati  negative
>>>

We also used the same data to train Swahili-Sentiment-Analysis model.

Issues

In case you're facing any difficulty when trying to use the data, please raise an issue so as we can assist you.

Contributions

Please Feel free to contribute to this repository, whether its code, docs or data processing pipeline.

Give it star

Did you find this repository useful, please give it a star so as more people can know about it

Credits

All the credits to

  1. Jinamizi
  2. Kalebu