Table 6.

The overall performance of i2b2 PHI sub-categories on the i2b2 test set at the strict entity level

Category	Sub-category	#Expected	#Predicted	#Correct	P	R	F
DATE	DATE	4,980	4,877	4,812	0.9867	0.9663	0.9764
NAME	DOCTOR	1,912	1,808	1,758	0.9723	0.9195	0.9452
	PATIENT	879	830	797	0.9602	0.9067	0.9327
	USERNAME	92	88	88	1.0000	0.9665	0.9778
LOCATION	HOSPITAL	875	807	727	0.9009	0.8309	0.8644
	CITY	260	214	184	0.8598	0.7077	0.7764
	STATE	190	167	154	0.9222	0.8105	0.8627
	STREET	136	134	132	0.9851	0.9706	0.9778
	ZIP	140	136	136	1.0000	0.9714	0.9855
	ORGANIZATION	82	35	25	0.7143	0.3049	0.4274
	COUNTRY	117	28	22	0.7857	0.1880	0.3034
	LOCATION-OTHER	13	1	0	0.0000	0.0000	0.0000
AGE	AGE	764	728	707	0.9712	0.9254	0.9477
ID	MEDICALRECORD	422	427	412	0.9649	0.9763	0.9706
	IDNUM	195	182	158	0.8681	0.8103	0.8382
	DEVICE	8	3	3	1.0000	0.3750	0.5455
CONTACT	PHONE	215	203	199	0.9803	0.9256	0.9522
	FAX	2	2	1	0.5000	0.5000	0.5000
	EMAIL	1	1	1	1.0000	1.0000	1.0000
PROFESSION	PROFESSION	179	132	107	0.8106	0.5978	0.6881