Table 2.

Classifiers used for automatic feature selection.

Classifier and text class			Accuracy		Macro average F1a		Precision		Recall		F1
Ridge classifier			0.925		0.89
	Adult-oriented readings					0.99		0.91		0.95
	Children-oriented readings					0.74		0.97		0.84
SVMb			0.93		0.89
	Adult-oriented readings					0.95		0.96		0.96
	Children-oriented readings					0.84		0.8		0.82
XGBc			0.94		0.90
	Adult-oriented readings					0.95		0.98		0.96
	Children-oriented readings					0.91		0.78		0.84

^aF1 = 2 × [(precision × recall) / (precision + recall)].

^bSVM: support vector machine.

^cXGB: extreme gradient boosting.