Research Article
Chinese Microblog Sentiment Detection Based on CNN-BiGRU and Multihead Attention Mechanism
Table 4
Comparison of test results with various traditional machine learning methods.
| ā | Accuracy | Precision | Recall | F1 |
| Decision tree | 0.7247 | 0.7277 | 0.7182 | 0.7229 | KNN | 0.8048 | 0.8290 | 0.7679 | 0.7973 | Naive Bayes | 0.8151 | 0.8252 | 0.7995 | 0.8122 | Random forest | 0.8455 | 0.8434 | 0.8485 | 0.8459 | GBDT | 0.8684 | 0.8640 | 0.8745 | 0.8692 | SVM | 0.8758 | 0.8682 | 0.8861 | 0.8771 | Logistic regression | 0.8890 | 0.8820 | 0.8981 | 0.8900 | CNN-BiGRU-MAttention | 0.9765 | 0.9901 | 0.9606 | 0.9751 |
|
|