Table 2.

Top reported system performance for studies predicting the gender of Twitter users using traditional machine learning (ML) methods. Result metrics are reflected in this table as reported in the original publications and are not necessarily comparable to each other.

Study	Language	ML method	Reported performance
			F1-score	Accuracy
Cesare et al [122], 2017	English	Ensemble: lexical match and SVMa and DTb	0.84	0.83
Jurgens et al [80], 2017	English	RFc ensemble	0.78	0.80
Ljubešić et al [85], 2017	Portuguese, French, Dutch, Spanish, German, and Italian	SVM	N/Ad	0.61-0.69
Markov et al [87], 2016	English, Spanish, Dutch, and Italian	LogRe	N/A	0.57-0.77
Mukherjee and Bala [92], 2016	English	NBf	0.75	0.71
Verhoeven et al [106], 2017	Slovenian	SVM	0.93	0.93
Volkova [110], 2015	English and Spanish	LogR	N/A	0.82
Xiang et al [116], 2017	English	SVM and PMEg	N/A	0.76
Cheng et al [65], 2018	English, Filipino, and Taglish	SVCh with lasso	0.84	0.84
Emmery et al [69], 2017	English	fastText	N/A	0.76
Giannakopoulos et al [72], 2018	N/A	SVM PNNi	N/A	0.87
Khandelwal et al [82], 2018	Code-mixed Hindi-English	SVM	N/A	0.9
Miura et al [89], 2018	Japanese	XGBoostj	N/A	0.89
van der Goot et al [104], 2018	English, Dutch, French, Portuguese, and Spanish	SVM	N/A	0.66-0.72
Alessandra et al [49], 2019	Italian	Ensemble: lexical match and SVM	N/A	0.75
Hirt et al [76], 2019	German	Ensemble: binary classifiers	0.81	N/A
Hussein et al [79], 2019	Dialect Egyptian Arabic	Ensemble: RF and LinRk	NA	0.77-0.88
Vicente et al [107], 2018	English and Portuguese	Ensemble: Face++, LinR, and SVM	N/A	0.93-0.97
Arafat et al [51], 2020	Indonesian	Multinomial NB	N/A	0.75
Baxevanakis et al [54], 2020	Greek	SVM	N/A	0.7
Garcia-Guzman et al [70], 2020	English	Bag of trees	0.64	0.64
López-Monroy et al [86], 2020	English and Spanish	Bag of trees	0.64	0.64
Pizarro [96], 2020	English and Spanish	SVM	0.82-0.84	N/A
Vashisth and Meehan [105], 2020	English	LogR	N/A	0.57
Wong et al [113], 2020	English	SVM	0.58-0.62	0.60

^aSVM: support vector machine.

^bDT: decision tree.

^cRF: random forest.

^dN/A: not applicable.

^eLogR: logistic regression.

^fNB: naive Bayes.

^gPME: projection matrix extraction.

^hSVC: support vector classifier.

ⁱPNN: probabilistic neural network.

^jXGBoost: extreme gradient boosting.

^kLinR: linear regression.