Table 2. Entity extraction evaluation based on manually annotated data.

	Human annotation, n	System annotation, n	Precision	Recall	F1-score
Female people	2,906	3,387	72.4%	77.6%	75.0%
Male people	8,381	10,034	77.4%	92.1%	84.2%
Female sources	1,442	1,104	94.6%	64.6%	76.8%
Male sources	3,809	3,346	87.7%	76.5%	81.8%