Table 10.

Aggregated results for the period 2010–2016

Case	Size	Precision	F-measure	Recall	Recall+
2010 - all	2103	0.68	0.791	0.944	0.852
2011 - all	4735	0.311	0.471	0.971	0.923
2012 - all	4114	0.359	0.525	0.975	0.934
2013 - all	4620	0.32	0.482	0.976	0.937
2014 - all	3271	0.448	0.613	0.968	0.914
2015 - all	2421	0.61	0.75	0.974	0.932
2016 - all	2445	0.611	0.754	0.985	0.96
2010 - top 3	1621	0.861	0.889	0.92	0.789
2011 - top 3	1590	0.892	0.913	0.935	0.831
2012 - top 3	1618	0.887	0.916	0.947	0.859
2013 - top 3	1645	0.884	0.921	0.96	0.894
2014 - top 3	1718	0.852	0.905	0.965	0.908
2015 - top 3	1738	0.842	0.899	0.965	0.908
2016 - top 3	1624	0.895	0.926	0.959	0.892
Union - best	1735	0.847	0.904	0.969	0.918
Union - all	10756	0.14	0.246	0.995	0.986

Best values for the different measures in the ‘top 3’ and ‘all’ categories are in bold face