Skip to main content
. 2021 Apr 14;7(4):e26527. doi: 10.2196/26527

Table 4.

Random forest model results. Random forest with active learning often, although not universally, shows improved performance compared to generic random forest models. The change between these two approaches is noted in the Change column.

Conspiracy theory and metrics Random forest Random forest with active learning Change
5Ga

Accuracy 0.779 0.783 0.004

Recall 0.908 0.872 –0.036

Precision 0.728 0.744 0.016

F1 Score 0.808 0.804 –0.004
Gatesb

Accuracy 0.622 0.5819 –0.04

Recall 0.675 0.793 0.118

Precision 0.608 0.556 -0.052

F1 Score 0.64 0.654 0.014
Labc

Accuracy 0.782 0.84 0.058

Recall 0.699 0.833 0.134

Precision 0.9 0.883 –0.017

F1 Score 0.787 0.857 0.070
Vaxd

Accuracy 0.507 0.751 0.244

Recall 0.653 0.474 –0.1786

Precision 0.170 0.274 0.104

F1 Score 0.270 0.347 0.077

a5G: conspiracy theories related to 5G technology.

bGates: conspiracy theories related to Bill Gates or the Bill & Melinda Gates Foundation.

cLab: conspiracy theories related to SARS-CoV-2 being laboratory-released or human-made.

dVax: conspiracy theories related to vaccines.