Table 7 Results of Ml models on the Urdu Tweet dataset using the CountVectorizer.
From: Ensemble stacked model for enhanced identification of sentiments from IMDB reviews
Measure | MNB(%) | BNB(%) | SVM(%) | DT(%) | RF(%) | LR(%) |
|---|---|---|---|---|---|---|
Classification accuracy | 57.99 | 56.99 | 52.00 | 53.0 | 55.00 | 56.99 |
Precision | 51.16 | 46.51 | 45.45 | 44.73 | 36.61 | 50.00 |
Recall | 57.16 | 51.00 | 58.13 | 39.63 | 30.23 | 53.48 |
F1 score | 53.99 | 48.65 | 51.02 | 41.97 | 33.01 | 51.15 |
Negative Recall | 63.15 | 60.00 | 47.36 | 63.15 | 73.64 | 59.64 |
