Sentimental

Python port of github.com/Wobot/Sentimental with some improvements

A simple dictionary-based sentiment analysis system with Russian language support.

Installation

pip install -U git+https://github.com/text-machine-lab/sentimental.git

or

pip install -U git+ssh://[email protected]/text-machine-lab/sentimental.git

Usage

from sentimental import Sentimental

sent = Sentimental()

sentence = 'Today is a good day!'
result = sent.analyze(sentence)

The result is a dictionary with four fields:

{'negative': 0.0, 'positive': 3.0, 'score': 3.0, 'comparative': 0.6}

The filed score reflects the overall sentiment of the input data, and the comparative field is normalized by the length of the input, so it can be used to compare the sentiment of different texts.

Citation

If you've found this project useful, please cite the following paper:

@inproceedings{rumshisky2017combining,
  title={Combining network and language indicators for tracking conflict intensity},
  author={Rumshisky, Anna and Gronas, Mikhail and Potash, Peter and Dubov, Mikhail and Romanov, Alexey and Kulshreshtha, Saurabh and Gribov, Alex},
  booktitle={International Conference on Social Informatics},
  pages={391--404},
  year={2017},
  organization={Springer}
}