Research Article
An Efficient Sentiment Classification Method with the Help of Neighbors and a Hybrid of RNN Models
Table 6
Cross-dataset evaluation results.
| Group | Method | Training dataset | Testing dataset | Accuracy | F1_score |
| One | LSTM | BG (all data) | BIS (train set) | 94.39 | 94.70 | BG (all data) | BIS (test set) | 94.94 | 95.35 | BG (all data) | BIS (all data) | 94.57 | 94.90 | GRU | BG (all data) | BIS (train set) | 94.66 | 94.94 | BG (all data) | BIS (test set) | 94.71 | 94.10 | BG (all data) | BIS (all data) | 94.67 | 94.97 | DTSC | BG (all data) | BIS (train set) | 95.96 | 96.23 | BG (all data) | BIS (test set) | 95.25 | 95.64 | BG (all data) | BIS (all data) | 95.92 | 96.21 |
| Two | LSTM | BG (train set) | BIS (train set) | 94.40 | 94.70 | BG (train set) | BIS (test set) | 92.89 | 93.25 | BG (train set) | BIS (all data) | 94.69 | 94.97 | GRU | BG (train set) | BIS (train set) | 93.78 | 94.05 | BG (train set) | BIS (test set) | 94.06 | 94.45 | BG (train set) | BIS (all data) | 93.21 | 93.46 | DTSC | BG (train set) | BIS (train set) | 95.50 | 95.94 | BG (train set) | BIS (test set) | 96.14 | 96.58 | BG (train set) | BIS (all data) | 95.68 | 96.11 |
|
|
Bold values indicate the best result.
|