Table 5. Experiment results of GloVe and OHE embedding models with deep learning algorithms for independent dataset.

The best accuracy percentages are represented in bold letters.

MODELS	ACC	SEN	SPE	MCC	AUC
GL+CNN	82.81	90.90	74.19	66.24	82.55
GL+LSTM	84.37	79.69	88.57	72.59	84.87
GL+BiLSTM	85.93	88.92	93.54	72.87	86.16
OHE+CNN	75.16	85.23	82.54	77.08	84.19
OHE+LSTM	57.81	58.36	67.46	51.17	60.09
OHE+BiLSTM	82.81	86.13	84.74	78.09	82.35