In this paper, we aim at investigating the binary classification problem for Vietnamese sentiment analysis, where our model can predict positive or negative sentiment given a user review
We demonstrate the power of our methods on two datasets:
Dataset | #Positive | #Negative | #Total | |
---|---|---|---|---|
AIVIVN Dataset | train | 8690 | 7383 | 16073 |
test | 5767 | 5214 | 10981 | |
Our Dataset | train | 11765 | 3235 | 15000 |
test | 137833 | 30210 | 168043 |
- Baseline methods:
- Our proposed ensemble methods:
@article{
}
Issues should be raised directly in the repository. For further support requests please email us at:
- Binh T. Nguyen ([email protected])