Research Article
Social Network Text Sentiment Analysis Method Based on CNN-BiGRU in Big Data Environment
Table 3
Experimental results of different dropout values.
| Dropout | Accuracy (%) | Precision (%) | Recall (%) | F1 (%) |
| 0.10 | 89.24 | 87.53 | 90.77 | 89.12 | 0.15 | 89.56 | 87.12 | 91.04 | 89.04 | 0.20 | 89.74 | 89.71 | 90.87 | 90.29 | 0.25 | 89.95 | 90.07 | 91.01 | 90.54 | 0.30 | 89.52 | 87.65 | 90.53 | 89.07 | 0.35 | 89.05 | 88.06 | 91.88 | 89.93 | 0.40 | 88.78 | 86.92 | 92.39 | 89.57 | 0.45 | 88.59 | 89.88 | 88.20 | 89.03 | 0.50 | 88.33 | 87.75 | 89.89 | 88.81 |
|
|
Bold marks the highest value.
|