Table 1.

Distributions of PHI categories in the training and test corpora

PHI Category	Sub-category	Training Data	Test Data
DATE	DATE	7,495	4,980
NAME	DOCTOR	2,877	1,912
	PATIENT	1,315	879
	USERNAME	264	92
AGE	AGE	1,233	764
CONTACT	PHONE	309	215
	FAX	8	2
	EMAIL	4	1
	URL	2	0
ID	MEDICALRECORD	611	422
	IDNUM	261	195
	DEVICE	7	8
	BIOID	1	0
	HEALTHPLAN	1	0
LOCATION	HOSPITAL	1,437	875
	CITY	394	260
	STATE	314	190
	STREET	216	136
	ZIP	212	140
	ORGANIZATION	124	82
	COUNTRY	66	117
	LOCATION-OTHER	4	13
PROFESSION	PROFESSION	234	179
Total		17,389	11,462