Table 5.

Transfer Learning Results for Model-2. First six rows for VHA test-set models and last six rows are for VCU test-set. Results reported are macro-averaged. Support indicated the total number of samples in test-sets. LM: Language Model. Bold values indicates the best results generated.

LM	Train	Test	Precision	Recall	F1-Score	Support
VHA	VHA	VHA	0.77	0.78	0.78	69
VHA	VCU	VHA	0.68	0.61	0.61	69
VCU	VHA	VHA	0.80	0.83	0.81	69
VCU	VCU	VHA	0.33	0.49	0.39	69
VHA_VCU	VHA	VHA	0.76	0.79	0.75	69
VHA_VCU	VCU	VHA	0.54	0.51	0.46	69
VHA	VHA	VCU	0.56	0.68	0.48	106
VHA	VCU	VCU	0.67	0.69	0.68	106
VCU	VHA	VCU	0.55	0.64	0.53	106
VCU	VCU	VCU	0.46	0.49	0.47	106
VHA_VCU	VHA	VCU	0.55	0.61	0.54	106
VHA_VCU	VCU	VCU	0.59	0.54	0.55	106