From: A deep learning-based model using hybrid feature extraction approach for consumer sentiment analysis
ABSA/SA Technique | Precision | Recall | F1-measure | Accuracy | AUC |
---|---|---|---|---|---|
SABSA | 0.845 | 0.832 | 0.839 | 0.829 | 0.816 |
SentiVec | 0.863 | 0.843 | 0.855 | 0.847 | 0.834 |
Ngram + TF-IDF + SVM | 0.851 | 0.838 | 0.845 | 0.837 | 0.823 |
SEML | 0.841 | 0.824 | 0.833 | 0.826 | 0.813 |
MTMVN | 0.793 | 0.773 | 0.785 | 0.777 | 0.765 |
HFV + LSTM | 0.946 | 0.912 | 0.922 | 0.902 | 0.918 |