A conversational agent system for dietary supplements use

Table 4 Performance of various question type classifiers

Model	Precision	Recall	F1	Accuracy (%)
Random Forest	0.603	0.556	0.540	55.60
SVM (with RBF Kernel)	0.629	0.629	0.620	62.90
SVM (with Linear Kernel)	0.647	0.649	0.640	64.90
Logistic Regression	0.656	0.649	0.64	64.90
BiGRU + Attention	0.737	0.728	0.729	72.85
LSTM	0.800	0.790	0.790	78.67
Bi-LSTM	0.800	0.790	0.790	79.41
CNN	0.810	0.800	0.810	80.79

ISSN: 1472-6947