You are viewing the site in preview mode

Skip to main content

Table 2 Data splits to train and test each model (listed in the header)

From: Hybrid natural language processing tool for semantic annotation of medical texts in Spanish

 

UMLS entities, temporal entities, miscellaneous entities, temporality/experiencer attributes

Medication information

Negation/speculation

 

#Texts

#Tokens

#Texts

#Tokens

#Texts

#Tokens

Train

720

175203

720

175203

7739

693271

Train (CC)

820

205011

1085

285876

7839

723079

Dev

240

58670

240

58670

240

58670

Test

240

58300

240

58300

240

58300

Test (HE)

200

27332

200

27332

200

27332

  1. CC: clinical cases; HE: human evaluation