Skip to main content
. 2022 Mar 25;52(14):16701–16717. doi: 10.1007/s10489-022-03342-5

Table 3.

Sentiment evaluation results

Evaluation methods BERT-wwm RoBERTa
Precision 0.83 0.81
Recall 0.84 0.79
F1_score 0.83 0.80