Table 4 Comparison of different sentiment lexicons.
From: Multi task opinion enhanced hybrid BERT model for mental health analysis
Lexicon | Task | Test accuracy (%) | Macro precision (%) | Recall (%) | F1-score (%) |
|---|---|---|---|---|---|
Afinn | Status classification | 91.35 | 90.53 | 92.09 | 91.23 |
Sentiment classification | 92.99 | 92.01 | 92.28 | 91.97 | |
VADER | Status classification | 92.27 | 91.84 | 92.94 | 92.36 |
Sentiment classification | 94.12 | 94.39 | 93.07 | 93.68 | |
TextBlob | Status classification | 93.74 | 94.17 | 94.48 | 94.32 |
Sentiment classification | 96.25 | 96.50 | 96.68 | 96.58 |
