Table 7 Comparison results of various evaluation metrics from multinomial Naïve Bayes with non-English Language stopwords removal.
From: Key insights into recommended SMS spam detection datasets
Multinomial Naïve Bayes | |||||||
|---|---|---|---|---|---|---|---|
Dataset | Accuracy | Precision (Ham) | Precision (Spam) | Recall (Ham) | Recall (Spam) | F1 - Score (Ham) | F1 - Score (Spam) |
2 | 98.28% | 0.99 | 0.98 | 0.97 | 0.99 | 0.98 | 0.98 |
4 | 90.10% | 0.89 | 0.93 | 0.95 | 0.84 | 0.92 | 0.88 |
6 | 96.75% | 0.96 | 0.97 | 0.97 | 0.96 | 0.97 | 0.97 |
7 | 92.95% | 0.94 | 0.92 | 0.92 | 0.94 | 0.93 | 0.93 |
8 | 96.94% | 0.99 | 0.95 | 0.94 | 0.99 | 0.97 | 0.97 |
9 | 90.48% | 0.80 | 1.00 | 1.00 | 0.85 | 0.89 | 0.92 |
10 | 86.49% | 0.80 | 0.91 | 0.86 | 0.87 | 0.83 | 0.89 |
