Table 1.

Descriptive statistics of data sets used in this study.

Data set		Notes	Entities	Entity types	Note types
i2b2 2010	Training	349	27,837	Problem, Treatment Test	Discharge Progress
	Test	477	45,009
SemEval 2014	Training	298	11,156	Disorder	Discharge, radiology ECG, ECHO
	Test	133	7,971
MIMIC II	N/A	403,871	N/A	N/A	Discharge, radiology ECG, ECHO

ECG: electrocardiogram, ECHO: echocardiogram