Table 6. Comparison of features used.
Performance comparisons.
| Techniques | Amazon 72K reviews dataset accuracy in past research (Krishnamoorthy, 2015) | Amazon (65K reviews) Accuracy on 12 features | Amazon (9.8 Mil Reviews) accuracy on 10 features |
|---|---|---|---|
| Support Vector Machine | 77.81% | 83.59% | 88.98% |
| Random Forest | 81.33% | 84.51% | 89.36% |
| Naïve Bayes | 71.27% | 81.68% | 84.75% |
| Decision Tree | — | 77.58% | 83.89% |
| Artificial Neural Network | — | 67.13% | 72.61% |
Notes.
The bold scores state the highest accuracy achieved.