Skip to main content
. 2020 May 30;2020:730–739.

Table 2.

Performance of tweet classification and gender identification models

Model Precision Recall F-score
Relevant vs. Irrelevant (1,397 vs. 442)
RF 0.73 0.66 0.68
CNN 0.75 0.76 0.76
LSTM 0.75 0.77 0.74
Promotional information vs. Consumers’ discussions (1,082 vs. 315)
RF 0.76 0.75 0.76
CNN 0.78 0.79 0.79
LSTM 0.77 0.76 0.76
Gender identification (male: 2,450 vs. female: 2,450)
RF 0.65 0.64 0.64
CNN 0.80 0.80 0.80
LSTM 0.58 0.58 0.57