Table 2.

Model performance in different training settings.

		No Pre + Raw		No Pre + BIO		Pre + Raw		Pre + BIO
		Best	Ave.	Best	Ave.	Best	Ave.	Best	Ave.
Test set (Span)	Precision	0.78	0.76	0.84	0.85	0.93	0.87	0.86	0.83
	Recall	0.62	0.63	0.68	0.66	0.80	0.70	0.71	0.7
	F1 score	0.69	0.66	0.75	0.74	0.89	0.78	0.78	0.73
Test set (Trunk)	Precision	0.54	0.52	0.58	0.53	0.74	0.61	0.63	0.63
	Recall	0.53	0.51	0.56	0.52	0.74	0.56	0.64	0.61
	F1 score	0.53	0.52	0.57	0.54	0.74	0.58	0.64	0.62