Table 1.
Splitting of datasets used in the experiments.
Dataset name | Total of samples | Training set (70%) | Validation set (10%) | Testing set (20%) | Total of deceptive reviews | Total of truthful reviews |
---|---|---|---|---|---|---|
Amazon | 21,000 | 15120 | 1680 | 4200 | 10500 | 10500 |
Yelp | 9460 | 6622 | 946 | 1892 | 4730 | 4730 |
Restaurants | 110 | 80 | 8 | 22 | 55 | 55 |
Hotels | 1600 | 1152 | 128 | 320 | 800 | 800 |