README.md

metadata

license: mit
task_categories:
  - text-classification
language:
  - vi
pretty_name: Multi-Level Sentiment Analysis
size_categories:
  - n<1K

Multi-level sentiment analysis of Products Reviews

• Size: 5971 reviews about products on e-commerce websites
• Labels: 5 level of senitments as follows:

Score	Sentiment
1	Very negative
2	Negative
3	Neutral
4	Positive
5	Very positive

• Source code: https://github.com/sonlam1102/multi-level-sentiments

Publication: (Please cite this publication when you use this dataset)

@inproceedings{fujita2021multi,
  title={Multi-Level Sentiment Analysis of Product Reviews Based on Grammar Rules},
  author={Hien D. Nguyen, Thanh Le, Khiem Tran, Son T. Luu, Suong N. Hoang, Hieu T. Phan},
  booktitle={New Trends in Intelligent Software Methodologies, Tools and Techniques: Proceedings of the 20th International Conference on New Trends in Intelligent Software Methodologies, Tools and Techniques (SoMeT\_21)},
  volume={337},
  pages={444},
  year={2021},
  organization={IOS Press}
}